High Court Ruling on Transfer Pricing Comparables and Bonus Deductions

Get the full picture - access the original judgement of the court order here

Case Name:

Chryscapital Inestment Advisors (India) Pvt. Ltd. Vs Deputy Commissioner of Income Tax (High Court of Delhi)

ITA 417/2014

Date: 27th April 2015

Key Takeaways:

1. High profit margins alone do not automatically disqualify a company from being a comparable for transfer pricing analysis.

2. The tax authorities must examine whether material differences in profitability can be reasonably adjusted under the law.

3. The assessee's claim for deduction of bonuses paid to shareholder-employees was allowed.

Issue:

1. Whether companies with "super profits" can be excluded as comparables for determining the arm's length price (ALP) of the assessee's international transactions.

2. Whether the assessee can claim deductions for bonuses paid to its shareholder-employees.

Facts:

- The assessee is a private investment advisory company that provides services to associated enterprises (AEs).

- For AY 2008-09, the assessee entered into international transactions with AEs amounting to over ₹56 crores.

- The assessee used the Transactional Net Margin Method (TNMM) to determine the ALP of these transactions.

- The tax authorities included certain companies with very high profit margins as comparables, which the assessee challenged.

- The tax authorities also disallowed the assessee's claim for deduction of bonuses paid to its two shareholder-employees.

Arguments:

- The assessee argued that companies with "super profits" should be excluded as comparables, as their high profitability distorts the ALP determination.

- The tax authorities argued that high profitability alone does not justify excluding a company as a comparable, unless there are material functional differences.

- The assessee contended that the bonuses paid to the shareholder-employees were based on their performance and qualifications, and should be allowed as deductions.

- The tax authorities claimed the bonuses were paid to avoid dividend distribution and should not be allowed as deductions.

Key Legal Precedents:

- Rule 10B of the Income Tax Rules, 1962 - Provides the methods and factors for determining ALP.

- Section 92C of the Income Tax Act, 1961 - Outlines the process for determining ALP.

- OECD Transfer Pricing Guidelines - Provide international guidance on comparability analysis and use of multiple year data.

Judgment:

- The High Court held that high profit margins alone do not automatically disqualify a company from being a comparable.

- The court directed the lower authorities to examine whether the material differences in profitability of certain comparables (Brescon, Khandwala Securities, Keynote) can be reasonably adjusted under Rule 10B(3).

- The court allowed the assessee's claim for deduction of bonuses paid to the shareholder-employees under Section 36(1)(ii) of the Act.

FAQs:

Q: Can companies with "super profits" be excluded as comparables for transfer pricing analysis?

A: No, high profit margins alone do not automatically disqualify a company as a comparable. The tax authorities must examine whether the material differences in profitability can be reasonably adjusted under the law.

Q: How should the tax authorities determine if a company with high profits can be used as a comparable?

A: The authorities must conduct an analysis under Rule 10B(3) to see if the material differences in profitability can be eliminated through reasonable adjustments. If the differences cannot be adjusted, then the company may be excluded as a comparable.

Q: Can the assessee claim deductions for bonuses paid to its shareholder-employees?

A: Yes, the High Court allowed the assessee's claim for deduction of bonuses paid to the shareholder-employees under Section 36(1)(ii) of the Income Tax Act, as the bonuses were based on their performance and qualifications.

Q: What is the significance of this case for transfer pricing analysis in India?

A: This case provides important guidance on the treatment of high-profit comparables and the use of multiple year data in determining the arm's length price. It emphasizes that the tax authorities must follow the specific provisions of the law and cannot automatically exclude companies based on their profitability.

1. Is there a concept of “super profit” in the arm‟s length price/transfer price determining process under the Income Tax Act, 1961 ("the Act") or the Rules framed thereunder, entitling tax administrators to include high profit making companies‟ data in the list of "comparables"? Benches of Income Tax Appellate Tribunal ("ITAT"), appear to be riven in their opinion on this; it is the subject matter of the present appeal.

2. The questions framed for decision in this appeal, under Section 260-A of the Act, arising from an order of the Income Tax Appellate Tribunal (“ITAT”) dated 20.12.2013 in ITA No. 6183/Del/2012 for assessment year (AY) 2008-09, are as follows:

1) Whether the proviso to Rule 10B(4) of the Income Tax Rules, 1962 will be applicable in case of fluctuations in the operating profit margins of comparable companies during the relevant financial year under question as compared to earlier years?

2) Whether comparables can be rejected on the ground that they have exceptionally high profit margins as compared to the assessee in transfer pricing analysis?

3) Whether factors like differential functional and risk profile coupled with high degree of volatility in operating profit margins is sufficient ground to reject comparables for transfer pricing analysis?

4) Whether disallowances can be made under Section 36(1)(ii) when the bonus paid to shareholders is not in the exact proportion of their shareholding and there is no avoidance of taxes?

3. The assessee is a private limited company incorporated under the Companies Act, 1956 and is engaged in providing investment advisory services, which were reimbursed on a cost-plus mark-up basis. Ashish Dhawan and Kunal Shroff, in the concerned assessment year, were its two shareholders - holding shares in the assessee in the proportion of 2:l; they were also its full time employees. In AY 2008-09, the assessee entered into international transactions with associated enterprises (AEs) relating to advisory services and reimbursement of expenses incurred on behalf of AEs amounting to ` 56,61,99,829/- and ` 4,49,72,912/- respectively. For the purposes of determination of arm‟s length price (ALP), the assessee used the Transactional Net Margin Method (“TNMM”). The assessee treated the transactions relating to reimbursement received by it from its associated enterprises on actual basis (i.e. without mark-up) at ALP as such since no value addition was done by it in relation to the said expenses. The assessee identified four entities which were engaged broadly in the same economic activities as in its case and identified as comparables. The result of the arm‟s length analysis is given below:

S. No. Comparable Entity

Operating Profit Margins 2005-06 2006-07 2007-08 Average

1. IDFC Investment Advisors Limited

- -55.50% 17.30% -19.10%

2. Future Capital Holdings Limited

- 0.88% 20.53% 10.71%

3. Khandwala Securities Limited

43.35% 42.62% - 42.99%

4. Sumedha Fiscal Services Limited

-16.47% -20.36% - -18.42%

Final Average 4.04%

4. The assessee‟s position was that because of fluctuation in the margins of the comparable entities, multiple year data of the comparables was warranted to remove the effect of year specific aberrations. Against the average Operating Profit Margin (“Operating Margin”) of 4.04% earned by the comparable entities, the assessee earned an Operating Margin of 27.05% and concluded that its transactions with its AEs were at arm‟s length. The assessee relied on Rule 10B(4) of the Income Tax Rules, 1962 (hereafter “the Rules”). Rule 10B(4) reads as follows:

―(4) The data to be used in analysing the comparability of an uncontrolled transaction with an international transaction or a specified domestic transaction shall be the data relating to the financial year in which the international transaction or the specified domestic transaction has been entered into:

Provided that data relating to a period not being more than two years prior to such financial year may also be considered if such data reveals facts which could have an influence on the determination of transfer prices in relation to the transactions being compared.‖

The assessee argued that using multiple year data is consistent with the OECD Guidelines as well as transfer pricing regulations of several developed jurisdictions. The Operating Margin of the assessee was stable in contrast to the comparable companies, described below:

Financial Year

Operating Margin

2005-06 24.15%

2006-07 21.14%

2007-08 27.05%

Average 24.11%

5. On 30.09.2008, the assessee filed its return for AY 2008-09 declaring a total income of ` 12,41,83,160. Its case was scrutinized by the AO who referred the matter to the Transfer Pricing Officer (“TPO”) under Section 92CA (3) of the Act. On 03.10.2011, the TPO passed an order recommending transfer pricing additions of ` 20,93,34,155/- to the income of the Assessee. The TPO computed the Operating Margins of the four comparables above using single year data i.e. for FY 2007-08 and ignoring the data for two prior financial years i.e. 2005-06 and 2006-07 while determining the ALP. The TPO concluded that multiple year data for the assessee‟s comparables could not be used but introduced two new comparables with abnormal business profits. The TPO also retained a comparable inspite of it showing abnormal growth in the assessment year under consideration and considered reimbursable expenses as part of operating expenses and corresponding reimbursement as part of operating revenue of the assessee for the purpose of determining the arm's length price. The TPO held that the assessee had not furnished any detail as to how the data for the earlier years had an impact on the profits in the concerned assessment year of the assessee or the comparables.

6. Based on the TPO‟s report, the AO passed the assessment order on 21.12.2011, confirming the recommendations of the TPO. The AO also disallowed the bonus paid by the assessee to its shareholder employees – M/s Ashish Dhawan (` 67,91,947) and Kunal Shroff (` 30,19,433) – under Section 36(l)(ii) of the Act. The assessee filed its objections against the draft assessment order before the Dispute Resolution Panel (“DRP”). The DRP, by order dated 21.09.2012, confirmed the transfer pricing additions as well as the disallowance of the bonus made by the Respondent. Thereafter, on 19.10.2012, the AO completed the assessment under section 143(3) read with section 144C of the Act assessing the income of the assessee after sustaining the transfer pricing additions made by the TPO and disallowing the bonus paid to its shareholders. The ITAT dismissed the Assessee‟s appeal by its order dated 20.12.2013 and confirmed the additions made by the Respondent.

7. All the lower authorities included three entities as comparables which had very high profit margins as compared with that of the assessee. These entities namely, Brescon Corporate Advisors Limited ("Brescon") (Operating Margin of 87.4%), Keynote Corporate Services Limited ("Keynote") (Operating Margin of 191.58%) and Khandwala Securities Limited ("Khandwala") (Operating Margin of 80.79%) had exceptional profit margins as compared with the Assessee (Operating Margin of 27.05%) and rejected three other comparables selected by the assessee (i.e. IDFC Investment Advisors Ltd. (17.35%), Sumedha Fiscal Services Limited (9.14%) and Future Capital Holdings Limited (20.56%). Khandwala had been selected as a comparable by the assessee itself based on the multiple year data for the comparability analysis. However, the TPO substituted the same with the data for the concerned financial year, in which Khandwala had exceptionally high profit margins. The ITAT upheld these findings and held that current year data should be used in the absence of abnormal or exceptional facts/circumstances in existence which could have an influence on the results as well as the determination of the transfer prices for the year under consideration. Further, the ITAT held that Rule 10B does not provide any basis to exclude an entity or eliminate it from the list of companies solely on the basis of high profitability. The authorities - including ITAT, held that the decisive factors for determining inclusion or exclusion of any entity in/from the list of comparables are the specific characteristics of the services provided by the said entities, assets employed, risks assumed, the contractual terms and conditions prevailing including the geographical location and size of the market, cost of labour and capital in the markets, etc. and high or low profit margins could not be criteria for inclusion or exclusion of entities in the list of comparables. Arguments of the assessee

8. The assessee submits that even if the ITAT‟s ruling on the issue is accepted, Brescon and Keynote should be excluded from the list of comparables as its (the assessee‟s) risk profile is not similar to that of those two companies. They are risk-taking entities whereas the assessee operates on a cost plus model wherein a guaranteed return of 25% on costs is assured to it. The assessee further argues that its functional profile is significantly different from that of Keynote. Unlike the assessee, Keynote is involved in capital market activities, including lead managing IPOs, Rights Offers, Buybacks and Takeovers. Also, Keynote considers its activities to be a Merchant Banker as evidenced by its Director's Report and Notes to Accounts of the concerned financial year. The assessee submits that in the audited financials of Keynote, there is no service-wise break-up of profits and therefore, the profitability of the advisory services segment (which maybe considered similar to the services being rendered by the assessee) is not available to be compared with the assessee‟s profitability.

The assessee argues that Keynote's profit margins have shown volatility over the years which could be attributed to abnormal business conditions and therefore Keynote should be rejected as a comparable altogether. The Operating Margins of Keynote for the last 5 years are as follows:-

Assessment Year Operating Margin

2004-05 (-)6.87%

2005-06 13.33%

2006-07 94.06%

2007-08 145.83%

2008-09 191.58%

9. The assessee highlights that CIT (Appeals) too had rejected Keynote in a preceding as well as succeeding assessment year i.e. AY 2007-08 and 2009-10. Further, Keynote has been excluded as a comparable by the DRP in a preceding assessment year i.e. AY 2006-07. In that order, dated 04.03.2013, the DRP observed:

―As regards choice of Keynote Corporate services as a comparable by TPO based on single year data, DRP finds no infirmity in principles. However, after analyzing the economic circumstances as highlighted by the assessee and corroborated from the annual report of the year, we do find it may not be a robust comparable.

According to the assessee ̳we would like to state that this company has very volatile profit margins and since the Ld. TPO has computed the ALP on the basis of single year data (data for the FY 2005-06 only) this company should not be included in the final set of comparable as it would lead to distortion of the ALP.‘ The assessee while determining the ALP considered data for three years which mitigated the high volatility in operating margins of this company. However on the basis of single year data the operating margins of this company will substantially inflate the operating margins.

―The volatility in the operating margin of this company is clearly evident from the three year profitability of the comparables submitted before you are the Ld. TPO vide submission dated May 18, 2009 (copy enclosed at page 139 of the paper book dated January 01, 2010 filed before the Hon‘ble Panel). The operating margin of this company during the FY 2003-04 was negative 6.87% and which converted to positive 13.33. In the FY 2004-05, thereby exhibiting the this margin further increased to 94.06% showing an even higher volatility (80 percent points) vis-à-vis previous year.‖

Further we would also like to state that Keynote can also not be considered a comparable to the assessee (on the basis of single year data) for the reason that on the basis of single year data this company is earning exceptionally high profits (i.e. 94%). It is further submitted that on the possible reasons due to which Keynote has derived exceptional profits during the year may be due to some alliances formed by it with some foreign companies during the year. The relevant extract (copy enclosed as Annexure

3) from the annual report of Keynote is given hereunder:

The company formed alliances with a Middle East based consulting company and with a Swiss based consulting company to offer its clients cross border transaction ability. Thus, the exceptional profit earned by Keynote during the relevant year may be due to such alliance formed by Keynote with other companies in Middle East and Swiss. The profit earned by it due to such alliance cannot be used for the arm‘s length analysis.‖

2.3.4 In view of the above reasons, the DRP directs TPO to exclude this comparable as it is not a robust comparable for this year. TP grounds are accordingly disposed off as above.‖

10. On the issue of disallowance of bonuses paid by the assessee to its two full-time shareholder employees, it is submitted that bonuses were paid to all its employees during the relevant financial year on the basis of their performance and qualifications. Both the individuals to whom the bonuses paid were disallowed have requisite qualifications, experience and expertise in the field of investment advisory services. Accordingly, keeping in view their experience, expertise and performance, the assessee had compensated them. The assessee submits that bonus under Section 36(1)(ii) of the Act is allowed as deduction if the same amount would not have been payable to the shareholders as profits or dividends if it had not been paid as bonus. The provision requires the sum paid as bonus to be exactly the same as to be payable as dividend in absence of the bonus for there to be a disallowance.

The assessee submits that the bonus paid to the shareholder employees is not in the same proportion as their shareholding. It is also submitted that the basis for disallowance of bonus paid – that no dividend was declared by the assessee – is incorrect as it paid interim dividend amounting to ` 5,47,47,000/- in the concerned assessment year. Thus, the bonus paid to the two shareholders was not in lieu of dividend and therefore, should be allowed as tax deductible expenditure.

11. Learned counsel argued that the ALP of an international transaction has to be determined by applying one of the methods provided in section 92- C (3) of the Act; it should be the most appropriate method and should also take into account prescribed factors. This is, counsel stated, elaborated in Rule 10-B of the Rules, which contemplates adjustment on account of functional and other differences. He contended that adopting of any method ultimately envisages comparison of like functions, transactions and enterprises. Rule 10B(2)(a) provides that specific characteristic of services rendered by the two entities should be compared in order to treat the same as comparables for the purpose of transfer pricing analysis. Counsel also referred to the OECD guidelines and argued that accurate ALP determination is dependent on flexibility and sound exercise of discretion. Chapter III of the OECD guidelines was relied on to say that they recommend that where can it be determined that some uncontrolled transactions have a lesser degree of comparability than others, they should be eliminated. He also referred to Section A-5 of OECD guidelines on ―selecting and rejecting potential comparables‖ and pointed out that as per para 3.56, wherever uncontrolled transactions have a lesser degree of comparability than others, they should be eliminated. Counsel stated that similarly, Para 3.57 states that if the range of comparables includes a sizeable number of observations, statistical tools that take account of central tendency to narrow the range (e.g. the inter-quartile range or other percentiles); Para 3.59 suggests that where the application of the most appropriate method produces a range of figures, a substantial deviation among points in that range may indicate that the data used in establishing some of the points may not be as reliable as the data used to establish the other points in the range or that the deviation may result from features of the comparable data that require adjustments.

12. Learned counsel also relied on A.7.3 of the OECD guidelines dealing with ―extreme results in the context of comparability considerations‖ to point out that extreme results might consist of losses or unusually high profits. These can affect the financial indicators that are looked at in the chosen method; some potential comparables have extreme results, further examination would be needed to probe such results. This important issue was overlooked by ITAT. Counsel relied on proviso to Rule 10-B (4) and stated that though the mandate of the law is ordinarily to rely upon comparables' data for the current year, in certain circumstances, it is possible for the authorities to rely on previous years' data restricted to two previous years. This is to eliminate any distorted picture which might be the consequence of adherence to the contemporaneous data, like in the present case.

13. It was argued that the DRP's order for AY 2006-07 had accepted the assessee's argument and excluded Keynote from the list of comparables, on the ground that the said concern had earned abnormally high or super profits. On that occasion, as compared with its previous year (AY 2005-06) profit level of 94%, the profit of the enterprise was 145%, registering a 51% increase over the previous year. This was considered to be too high to be allowed as a comparable. During the current year, the profit registered was 191%. In the circumstances, it was illogical and arbitrary for the revenue to have rejected the contention that data in respect of Keynote should have been excluded. It was also similarly argued that the ITAT fell into error in rejecting the assessee's objection with respect to Brescon whose total turnover was over ` 14 crores, of which the comparable business was only ` 2 crores; the absence of any sectional data with regard to this company, meant that its activities were not comparable, on a fair application of Rule 10-B (2) and (3).

14. Learned counsel relied on the decisions of the Special Bench in the case of Quark Systems Private Limited v. DCIT (2010 38 SOT 307- Chandigarh Bench) Adobe Systems India Pvt. Ltd. (Del) 2011-(TII)-13- ITAT-DEL); Teva India (P) Ltd v. DCIT, [2011] 44 SOT 105 (Mum); Sapient Corporation (P) Ltd. v. Deputy CIT, [2011] 11 Taxmann 69 (Delhi); Asst CIT vs. Maersk Global Services Centre (India) P. Ltd. (133 ITD 543)(Mum.); Symantec Software Solutions (P) Ltd. v. Assistant CIT [2012] 25 Taxmann 163 (Mum); and a Division Bench decision of this court, in Commissioner of Income Tax v Agnity India Technologies Pvt. Ltd. (2013) 219 Taxman 26 (Del), were relied on. In Agnity India (supra) it was held that huge turnover companies like Infosys and Wipro cannot be considered as comparable to smaller companies like assessee.

15. Learned counsel for the assessee also argued that the rejection of previous years‟ data, in the facts of the present case, was unwarranted. It was submitted that given that the comparables introduced by the TPO distorted the margins, the AO and DRP erred in determining the ALP on the basis of data for financial year 2007-08 only and ignoring the data for two prior financial years i.e. FY 2005-06 and FY 2006-07. Learned counsel submitted that the TPO had the option of reaching back to previous years‟ data, since such power exists by virtue of proviso to Rule 10B (4). Learned counsel also relied on Part B.3, Paras 3.75 to 3.78 of OECD guidelines, in support of the submission.

Revenue‘s contentions

16. Mr. Rohit Madan, learned counsel for the revenue argued that five methods have been prescribed to determine ALP in relation to an international transaction and the comparability analysis requirements are method specific under Rule 10-B (1). Referring to the said Rule it was submitted that price charged or paid for the property transferred or service rendered in the comparable transaction is relevant in case of CUP and re- sale price method while the cost of production incurred in respect of property transferred or services provided is relevant for cost plus method. However, there is no mention of any property transferred or services provided in case of TNMM. They are provided for other methods. He contended that the relevant Rule thus makes it clear that specific characterization of the property transferred or services is not relevant for TNMM and this position is in conformity with the relevant OECD guidelines which suggest that broad comparability of functions should be done for TNMM.

17. Countering the submissions of the assessee, it was argued that neither the Act, nor the Rule contemplate exclusion of relevant transactions of like enterprises, in any manner other than what is prescribed. It was argued here that a comparable cannot be removed from consideration merely because it suffers loss; likewise, a unit or enterprise which enjoys higher profit (than the assessee or a significantly high profit in the industry) or even one making a so called "super profit" too cannot be eliminated. Generally, both loss making units and high profit making units cannot be removed from the list of comparables unless, such removal is statutorily permitted by Rule 10-B (2) or (3). Counsel also submitted that this is also evident from a reading of Rule 10-C. It was pointed out that Rule 10B (3) (ii) and Rule 10 C (2)(e) permitted adjustment to eliminate material defects of the difference between the assessee and comparables. Counsel argued that only those factors which result in material difference in the comparables of transactions as between the assessee and the unrelated transaction or the third party enterprise, have to be reasonably adjusted to avoid distortions under the said provisions. The step envisioned there had to be necessarily followed keeping in view the mandate "shall".

18. It was also argued that the decision in Commissioner Of Income Tax v Mentor Graphics (Noida) Pvt.Ltd (ITA 1114/2008, decided by this court on 04-04-2013) has held that OECD guidelines cannot be applied because there are specific provisions of Rule 10B (2) & (3) and the first proviso to Section 92C(2) which apply. There, it was held that having held that the comparables given by the assessee were to be accepted and those searched by the TPO were to be rejected, the only option then left to the ITAT was to derive the arithmetical mean of the profit level indicators of the comparables. It was submitted that accepting the theory of "abnormally high profits" as a ground for rejection of a comparable would lead to vagueness and confusion because what constitutes abnormally high has nowhere been spelt out in the Act or rules. On the other hand, the margin of variation permitted is ± 3% (proviso to Section 92C (2), reduced from the 5% margin that existed earlier). Introduction of any other variation not based in law would not be justified.

Analysis & Conclusions

19. Section 92-C which is relevant, for the purpose of determining ALP inter alia, reads as follows:

"92C. (1) The arm's length price in relation to an international transaction [or specified domestic transaction] shall be determined by any of the following methods, being the most appropriate method, having regard to the nature of transaction or class of transaction or class of associated persons or functions performed by such persons or such other relevant factors as the Board may prescribe, namely :-

(a)comparable uncontrolled price method;

(b) resale price method;

(c)cost plus method;

(d)profit split method;

(e)transactional net margin method;

(f)such other method as may be prescribed by the Board.

(2) The most appropriate method referred to in sub-section (1) shall be applied, for determination of arm's length price, in the manner as may be prescribed:

Provided that where more than one price is determined by the most appropriate method, the arm's length price shall be taken to be the arithmetical mean of such prices:

(3) Where during the course of any proceeding for the assessment of income, the Assessing Officer is, on the basis of material or information or document in his possession, of the opinion that-

(a) the price charged or paid in an international transaction [or specified domestic transaction] has not been determined in accordance with sub-sections (1) and (2); or

(b)any information and document relating to an international transaction [or specified domestic transaction] have not been kept and maintained by the assessee in accordance with the provisions contained in sub-section (1) of section 92D and the rules made in this behalf; or

(c)the information or data used in computation of the arm's length price is not reliable or correct; or

(d)the assessee has failed to furnish, within the specified time, any information or document which he was required to furnish by a notice issued under sub-section (3) of section 92D, the Assessing Officer may proceed to determine the arm's length price in relation to the said international transaction [or specified domestic transaction] in accordance with sub-sections (1) and (2), on the basis of such material or information or document available with him:

Provided that an opportunity shall be given by the Assessing Officer by serving a notice calling upon the assessee to show cause, on a date and time to be specified in the notice, why the arm's length price should not be so determined on the basis of material or information or document in the possession of the Assessing Officer."

20. Section 92C(1) thus visualizes determination of the “arms-length price” (ALP) by any of five enumerated methods, ―being the most appropriate method‖, having regard to the ―nature of transaction or class of transaction or class of associated persons or functions performed by such persons or such other relevant factors as the board may prescribe, namely (a) comparable uncontrolled price method, (b) resale price method, (c) cost + method, (d) profit split method, (e) transactional net margin method, (f) any such other method as may be prescribed by the board. Where more than one price is determined by the most appropriate method, the arm's length price shall be taken to be arithmetical mean of such prices.‖

21. Rule 10B of the Rules prescribes the determination of arm's length price under Section 92C. The first step in all methods is evaluation of differences between the international transaction undertaken with the ―unrelated enterprise performing the comparable functions‖ in similar circumstances. Rule 10B of the Income-tax Rules inter alia, provides for various methods for determination of the arm's length price. Rule 10B (1) (e) prescribes the “transactional net margin method” (TNMM) with which the present case is concerned. Rule 10B (1) (e) (i) is as under: "10B. (1) Determination of arm's length price under section 92C

(e) transactional net margin method, by which,—

(i) the net profit margin realised by the enterprise from an international transaction entered into with an associated enterprise is computed in relation to costs incurred or sales effected or assets employed or to be employed by the enterprise or having regard to any other relevant base."

22. These provisions prescribe, therefore, that even under the TNMM, importance is given to "assets employed or to be employed" as relevant factors for consideration. Rule 10B (2), as the second step, requires application of functions, asset, risk test for judging comparability of international transaction with an uncontrolled transaction. It provides:

"10B (2). For the purposes of sub-rule (1), the comparability of an international transaction with an uncontrolled transaction shall be judged with reference to the following, namely :—

(a) the specific characteristics of the property transferred or services provided in either transaction ;

(b) the functions performed, taking into account assets employed or to be employed and the risks assumed, by the respective parties to the transactions ;

(c) the contractual terms (whether or not such terms are formal or in writing) of the transactions which lay down explicitly or implicitly how the responsibilities, risks and benefits are to be divided between the respective parties to the transactions ;

(d) conditions prevailing in the markets in which the respective parties to the transactions operate, including the geographical location and size of the markets, the laws and the Government orders in force, costs of labour and capital in the markets, overall economic development and level of competition and whether the markets are wholesale or retail.

(e) the extent to which reliable and accurate adjustments can be made to account for differences, if any, between the international transaction or the specified domestic transaction and the comparable uncontrolled transaction or between the enterprises entering into such transactions;

(f) the nature, extent and reliability of assumptions required to be made in application of a method."

Rule 10B (3) stipulates the third step, and spells out when the TPO is obliged to hold an uncontrolled transaction as comparable with others. This provision reads as follows:

―(3) An uncontrolled transaction shall be comparable to an international transaction or a specified domestic transaction if-

(i) none of the differences, if any, between the transactions being compared, or between the enterprises entering into such transactions are likely to materially affect the price or cost charged or paid in, or the profit arising from, such transactions in the open market; or

(ii) reasonably accurate adjustments can be made to eliminate the material effects of such differences.‖

Rule 10B (4) provides what should be the basis of the calculations in terms of data, its contemporaneity, etc. It stipulates that:

―(4) The data to be used in analysing the comparability of an uncontrolled transaction with an international transaction shall be the data relating to the financial year in which the international transaction has been entered into:

Provided that data relating to a period not being more than two years prior to such financial year may also be considered if such data reveals facts which could have an influence on the determination of transfer prices in relation to the transactions being compared.‖

23. The assessee's argument is that entities earning “super normal” or “abnormal” profits should be excluded from the list of comparables. For this purpose, it relied on several rulings of various Benches of the ITAT. These are Adobe Systems India (P) Ltd. v. Additional Commissioner of Income-tax, [2011] 44 SOT 49 (Delhi) Teva India (P) Ltd v. DCIT, [2011] 44 SOT 105 (Mum); Sapient Corporation (P) Ltd. v. Deputy CIT, [2011] 11 Taxmann 69 (Delhi); Asst CIT vs. Maersk Global Services Centre (India) P. Ltd. (133 ITD 543)(Mum.); Symantec Software Solutions (P) Ltd. v. Assistant CIT [2012] 25 Taxmann 163 (Mum); and a Division Bench ruling of this court in CIT v. Agnity India Technologies (P) Ltd. [2013] 36 Taxmann 289 (Del HC). Besides, this court notices that a similar reasoning – of applying what is known as the “turnover” filter or the exclusion of “superprofit” making companies reasoning was applied in Continuous Computing India (P) Ltd. vs. ITO (2012) (52 SOT 45)(Bang)(URO); Centillium India P. Ltd vs. DCIT (2012)(20 ITR 69) (Bang)(Tri.) and Addl CIT vs. Frost and Sullivan India (P) Ltd (supra). The revenue has on the other hand, relied on contrary views in Actis Advisers P. Ltd. v. Deputy CIT [2012] 20 ITR (Trib) 138 (Delhi); 24/7 Customer.Com.Pvt.Ltd. v. Deputy CIT [2013] 21 ITR (Trib) 514 (Bang) and Willis Processing Services (I) P. Ltd. v. Deputy CIT [2014] 30 ITR (Trib) 39 (Mum). Such views are echoed in Trilogy E-Business Software India P. Ltd. v. Deputy CIT [2013] 23 ITR (Trib) 464 (Bang) and Stream International Services P. Ltd. v. Asst. DIT (International Taxation) [2013] 23 ITR (Trib) 70 (Mum) too.

24. Before analysing the relative strengths of the rival contentions, a tabular statement containing the reasoning which persuaded various Benches of the ITAT to conclude one way or the other is reproduced below:

S. No. Judgment Finding Rationale

1. ITO v. Saunay Jewels (P) Ltd., [2010] 42 SOT 2 (Mum).

1. One of the four comparables chosen by the TPO (Sovereign Diamonds Ltd.) should be excluded.

2. Simple arithmetic average of gross profit margin cannot be adopted as there is a wide variation in

The excluded comparable had a gross profit margin of 53.81% which was abnormal profits. the parameters. Weighted average should be taken.

2. Adobe Systems India (P) Ltd. v. Additional Commissioner of Income-tax, [2011] 44 SOT 49 (Delhi)

1. Directed the exclusion of three entities as comparables.

The said entities had shown supernormal profits. By excluding these three companies, the arithmetic mean of OP/TC comes to 17.15%, which falls within the +-5% range as permitted by s. 92C(2).

Further, the DRP has passed a cursory order without examining the submissions of the assessee.

3. Teva India (P) Ltd v. DCIT, [2011] 44 SOT 105 (Mum).

Remitted the matter to the AO to decide the issue of inclusion of M/s Vimta Labs as a comparable

M/s Vimta Labs had earned supernormal profits. The ITAT noted the decision in Adobe Systems and directed the matter to be decided in light of that decision and taking into account the submissions of the assessee.

4. Sapient Corporation (P) Ltd. v. Deputy CIT, [2011] 11 Taxmann 69 (Delhi)

Directed the exclusion of one of the comparables considered by the TPO (Zenith Infotech Ltd.)

TPO cannot exclude all loss making comparables and include an entity (Zenith) making supernormal profits at the same time.

Zenith is predominantly a software product company whereas the assessee is a software development services company and a software product company shows higher margin.

5. Nortel Networks India (P) Ltd. v. Additional CIT, [2013] 36 Taxmann 439

Affirmed the exclusion of M/sArraycom as a comparable.

A concern will not lose its status merely because it is a loss- making entity. However, TPO has not excluded Arraycom for the sole reason that it is a loss- (Delhi)

Further held that the TPO has adequately factored the subjective elements in determining the ALP.

making entity but because it has been showing persistent losses. Its operation also has a reducing tendency.

In the absence of exceptional circumstances, previous year data under the proviso to Rule 10B(4) cannot be used.

6. Carlyle India Advisors (P) Ltd. v. Additional CIT, ITA No. 7901/Mum/2011 dated 04/04/2012.

Directed the exclusion of, inter alia, Keynote Corporate Services Ltd. and S.R.E.I Capital Markets Ltd. as comparable

Keynote was into merchant banking whereas the assessee provided investment advisory and related support services.

SREI Caps‘ core business was merchant banking and consultancy income accounted for only 0.27% of the total income.

Absence of segmental data insofar as the investment advisory service provided by the assessee is concerned led to the exclusion of comparables.

7. Deputy CIT v. DeloitteConsulting India Pvt. Ltd., ITA No. 1082/Hyd/2010 dated 22/07/2011

Inclusion of Vishal Information Technology Limited as a comparable was not incorrect.

Wipro cannot be a comparable.

Assessee derived its income from software development and IT enabled services.

Assessee itself argued before the TPO that VTIL is a comparable company offering IT enabled services.

The intangibles will not Previous year data can be used for comparables only under exceptional circumstances.

materially affect the price or profit-earning [within the meaning of Rule 10B(3)]. No two comparable companies can be replicas of each other. Rule 10B should be applied on a broader perspective and not with technical rigour. Wipro cannot be a comparable as its turnover is 20 times that of the assessee.

8. Symantec Software Solutions (P) Ltd. v. Assistant CIT, [2012] 25 Taxmann 163 (Mum).

Two entities (ICC International Agricultural Ltd. and TSR Darashaw Ltd.) were directed to be excluded as comparables.

These entities were required to be excluded on account of significantly higher operating margins (82.92% and 78.29%) whereas the next highest was 26.67%. Thus, unless it was demonstrated that these super normal profits were earned in the normal routine of activities, they could not be included.

9. Sony India (P) Ltd. v. Deputy CIT, [2008] 114 ITD 448 (Delhi)

Upheld the revenue‘s decision to exclude Godrej as a comparable.

Reversed the revenue‘s finding on inclusion on Videocon as a comparable.

Exclusion may not be justified on the mere ground of loss and competition. However, on the facts of the case, a number of factors have the cumulative effect of justifying Godrej‘s exclusion. These are: Godrej makes refrigerators and not TVs, it has suffered huge losses over a period of several years, had huge unutilized capacity, needs financial restructuring, joint venture of the company stands terminated, etc.

Re inclusion of Videocon, there are material differences which cannot be eliminated within the meaning of Rule 10B(3). Thus, Videocon has to be excluded as a comparable.

10. Philips Software Centre v. ACIT, [2008] 26 SOT 226 (Bang.)

Companies with supernormal profits should have been excluded from the list of comparables by the TPO.

An entity making supernormal profits cannot be a comparable. If at all it were to be considered as a comparable, appropriate adjustments to the material differences would have to be made. However, normalization of the margins of super profit making companies is not envisaged on an ad hoc basis and has to be done as per the law.

The assessee was a captive contract service provider and it did not bear any business and operational risks

11. E-gain Communication (P) Ltd. v. ITO, [2008] 23 SOT 385 (Pune)

Excluded Thirdware Solutions Ltd. and WTI Advanced Technology as comparables.

The margin of profit shown by these two entities was extraordinary.

All factors materially affecting the comparability of the assessee with the other entities need to be scrutinized and adjusted, including the operative profit.

12. SAP LABS India (P) Ltd. v. ACIT, [2011] 44 SOT 156 (Bang.)

Directed the exclusion of M/s Hinduja TMT and M/s Aftek Infosys Ltd. as comparables.

These two entities were earning supernormal profits. Extreme cases should be avoided while making a comparative study of analogous cases.

13. Exxon Mobil Company India P. Ltd. v. Deputy CIT [2012] 15 ITR (Trib) 353 (Mum)

Rejected the assessee‘s contention that two loss making concerns had to be included among comparables.

As regards exclusion of entities earning abnormal profits, a general submission cannot be accepted; the assessee should bring out the peculiar features why such exclusion is necessary in the circumstances of the case.

14. Maersk Global Centres (India) (P) Ltd. v. ACIT, [2014] 43 Taxmann 100 (Mumbai Special Bench).

Entities with abnormally high profit margins cannot be rejected outright ascomparables. In the given facts of the case, two comparables sought to be included indicated unusual features for the year, which qualified for their exclusion.

The inclusion of entities with supernormal profits would depend upon the facts and circumstances of each case. It should trigger further investigation to establish whether it can be taken as a comparable or not – this would depend upon whether the high profits reflect a normal business condition or whether they are a result of some abnormal conditions prevailing in the relevant year.

The profit margin earned by such entity in the immediately preceding year may also be taken into account to determine this issue.

If the high profit margin does not reflect normal business condition, it should be rejected.

An entity cannot be rejected solely on the basis of abnormally high profit margin.

15. Goldman Sachs (India) Securities Pvt. Ltd. v. ACIT, ITA No. 7724/Mum/2011,

Directed the exclusion of comparables ordered by the TPO.

Assessee and the comparables were functionally different and not in the same segment. dated 23.01.2013

16. Advance Power Display Systems Ltd. v. ACIT, [2013] 35 Taxmann 145 (Mum)

Directed the exclusion of BCC Fuba India Ltd. as a comparable.

Comparables have to be tested for each year independently. The fact that an entity has been chosen as a comparable for one year does not ipso facto mean that it would be chosen the subsequent year.

BCC Fuba India Ltd. was a persistently loss making unit and therefore, it cannot be considered to be a good comparable.

Further, in respect of another company, the P&L A/c had an extraordinary item of income on account of sale of business. Therefore, this makes this company as not a good comparable for the year under consideration.

17. Syscom Corporation Ltd. v. ACIT, [2013] 35 Taxmann 600 (Mum)

A company cannot be excluded as a comparable solely because it is a high profit making unit.

A persistently loss making unit cannot be considered as a comparable.

Comparability of an uncontrolled transaction with an international transaction has to be measured by using current year data and only when the

If profit not supernormal, the mere fact that it is high does not justify exclusion. Unless and until there are specific reasons and factors as provided under Rule 10B, an entity cannot be excluded or eliminated from the list of comparables.

current year data does not give a true picture due to abnormal circumstances that multiple year data is used.

18. Aztec Software v. ACIT, ITA No. 584/Bang/2006 dt. 12.07.2007 (Bang – Special Bench)

The criteria prescribed under the Act and Rules is the primary basis for testing comparability.

There should be a proper analysis of transactions – FAR analysis.

19. CIT v. Agnity India Technologies (P) Ltd. [2013] 36 Taxmann 289 (Delhi HC)

Upheld the exclusion of Infosys Technologies Ltd. as a comparable.

Tribunal had excluded Infosys as it was a giant company in the area of software development and it assumed all risks leading to higher profits whereas the assessee was a captive unit of the parent company and assumed only a small risk. HC upheld the reasons given by the Tribunal for the exclusion.

20. Cummins Turbo Technologies v. DDIT, [2013] 35 Taxmann 350

Companies with supernormal profits and companies which are loss-making cannot straight away be rejected as comparables unless abnormal loss is projected.

21. Google India (P) Ltd. v. DCIT, [2013] 29 Taxmann 412.

Exclusion of two companies making supernormal profits.

The Tribunal has consistently held that super profit making companies have to excluded from the list of comparables before making transfer pricing adjustment.

25. Maersk Global Centres (India) (P) Ltd (supra) was a Special Bench (3 Member) decision of ITAT which had to address the precise question which arises for consideration in this case, i.e whether in the facts of that case ―companies earning abnormally high profit margin should be included in the list of comparable cases for the purpose of determining the arm's length price of international transactions‖. Although the ITAT did not specifically answer the question, in view of its findings that two comparables, i.e eClerx Services Ltd and Mold Tech Technologies Ltd, on account of unusual or peculiar features which were apparent from the materials on record, the Bench did indicate the general approach appropriate in this regard:

―the comparability of an international transaction with an uncontrolled transaction for the purpose of determining the arm's length price of an international transaction by following the transactional net margin method is required to be judged with reference to the functions performed as per sub-rule (2)(b) of rule 10B read with sub-rule (1)(e) thereof and there is no bar in the transfer pricing regulations in India to exclude certain entities selected as potential comparables on a broad functionality test by applying the functional test at narrow or micro level to attain the relatively equal degree of comparability. On the other hand, rule 10B(3) provides that the uncontrolled transaction selected/judged as per rule 10B(2) shall be comparable to an international transaction only if none of the differences, if any, between the transactions being compared, or between enterprises entering into such transactions are likely to materially affect the price or cost charged or paid or the profit arising from such transaction in the open market or reasonably accurate adjustment can be made to eliminate the effects of such difference. In our opinion, sub-rule (3) of rule 10B thus clearly provides for further exclusion of the comparables selected by applying the test/criteria given in sub-rule (2) of rule 10B if there is any difference found between the enterprises entering into the transactions which materially affects the cost charged or the profit arising from such transaction in the open market.

69. Keeping in view the relevant portion of the OECD Transfer Pricing Guidelines discussed above and having regard to the relevant transfer pricing regulations as contained in rule 10B(3) of the Income-tax Rules, 1962, we are of the view that further dissection or classification of information technology enabled services can be done depending on the facts and circumstances of each case so as to select the entities having a relatively equal degree of comparability.‖

In Exxon Mobil Company India P. Ltd. (supra), a Mumbai Bench decision, (cited at Sl. No.13 in the table above), the ITAT held:

"(xi) Now, coming to the alternative arguments of the assessee that abnormal profit making unit is also to be eliminated on the same analogy on which loss making units are excluded, we, in principle, do not dispute this proposition. The various case laws relied upon by the assessee lay down that a comparable cannot be eliminated just because it is a loss making unit. Similarly, a higher profit making unit cannot also be automatically eliminated just because the comparable company earned higher profits than the average. The reason for rejecting the two loss making units is not just because they were loss making units but for the reasons which are already stated in the preceding paragraphs. If similar reasons existed in the higher profit making unit, then, it is for the assessee to bring out those reasons and seek exclusion of the same. A general argument that you have to exclude units which have high profit range, in case you exclude units which have made loss is a general submission which cannot be accepted. In other words, as a general principle, both loss making unit and high profit making unit cannot be eliminated from the comparables unless there are specific reasons for eliminating the same which is other than the general reason that a comparable has incurred loss or has made abnormal profits.‖

This court notices that American Express Services India Ltd v Deputy Commissioner Of Income-Tax, 2013 (57) SOT 22 (ITAT-Del) said, similarly, that:

―If the comparables are performing the same functions then merely on the ground of they being earning super profits, cannot be excluded. Material differences between their business modules, however, are required to be taken care off and duly adjusted. In the case of Sundaram Finance Distribution Ltd., we find that the main objection of assessee is that the said comparable was included because assessee had supplied the same and the second objection is that in the said comparable there was no staff. As far as first objection is concerned, we are in agreement with the assessee's counsel that merely because the said comparable was provided by assessee, the same could not be included without proper examination to account for the differences. The assessee is well within his right to demonstrate that a comparable supplied by it in the transfer pricing analysis was not correct and had to be excluded. This right of the assessee is not curtailed in any manner, whatsoever, in the rules. A similar reasoning was adopted in M/s. Premier Exploration Services Pvt.Ltd., vs. ITO, Ward 14 (3) [2014] 29 ITR (Trib) 427 (ITAT) [Del]

―Although assessee had taken this company as comparable on the basis of past years data but in our considered view, the Saket Projects Ltd. was not comparable to assessee because the event management was done by sponsorships which is evident from various documents placed in paper book. Further the segment allocation of expenses also appears to be not reliable. We agree with the view of revenue that no comparable can be rejected merely on the basis of high margins if the comparable is functionally comparable to the assessee and also that there is miner variation in functional similarity. However, in the case of Saket Projects Ltd. there is functional dissimilarity. The company is organizing events with various kinds of sponsorships. The facts also suggest that segmental allocation of expenses were not reliable. We also hold that when direct comparables are available then segmental results of companies engaged in other business should not be taken as comparable. On the basis of these facts, we hold that Saket Projects Ltd. was not comparable to the extent wherein the various variations could be ruled out or iron out by provisions of law and rules.‖

26. The assessee‟s position is supported by reasoning in cases like the ITAT‟s decision in Mentor Graphics (Noida) (P.) Ltd. v. Dy. CIT [2007] 109 ITD 161 where contentions such as these were accepted:

".....The wide difference in the ratio of operating margins in the final selection of comparable ... is a clear pointer to the fact that the selection made was faulty...The OECD guideline on this point is as under

̳1.47 Where the application of one or more methods produces a range of figures, a substantial deviation among points in that range may indicate that the data used in establishing the some of the points may not be as reliable as the data used to establish the other points in the range or that the deviation may result from features of the comparable data that require adjustments.‘ Inferring from the above ruling, we requests your goodself to not consider companies displaying abnormal profits since they deviate from the normal trend displayed by the data set." Many decisions of different benches of the ITAT indicate a rote repetition (in the words of Felix Frankfurter J, quoted in the beginning of this judgment a "lazy repetition") of this reasoning, without an independent analysis of the provisions of the Act and the rules. (Ref. IQ Information Systems India P. Ltd. [2013] 25 ITR (Trib) 185 (Hyderabad Bench) Symphony Marketing Solutions India P. Ltd. [2013] 27 ITR (Trib) 753 (Bangalore Bench)).

27. An indication of what ought to be the correct approach was given by a Division Bench of this Court in Commissioner of Income Tax v Mentor Graphics (P) Ltd [2013] 259 CTR 1 (Del), where it was held that:

―21. The sum and substance of the Tribunal's order is that the criteria adopted by the TPO for searching comparables was not correct. Secondly, the TPO had not specifically rejected any of the comparables of the respondent/assessee. The Tribunal was of the view that the comparables of the respondent/assessee ought to have been accepted and, had that been the case, there would have been no need for the TPO to search for comparables. Of course, in passing the order, the Tribunal made certain general observations that unless and until the comparables drawn by the taxpayer were rejected, a fresh search by the TPO could not be conducted. However, this has to be tempered with the relevant statutory provisions which are clearly set out in sub-s. (3) of s. 92C of the said Act which stipulates four situations whereunder the AO/TPO may proceed to determine the ALP in relation to an international transaction. If any one of those four conditions is satisfied, it would be open to the AO/TPO to proceed to determine the ALP. This clarification of the observation of the Tribunal was necessary and that is why we have done so.

22. We also note that the Tribunal had gone further and reduced the list of comparables to merely four as indicated in para 46 of the impugned order. We do not think that it was the right approach to be adopted by the Tribunal. The Tribunal should have stopped at the point where it decided on facts that the comparables given by the respondent/assessee were to be accepted and those searched by the TPO were to be rejected. The only option then left to the Tribunal was to derive the arithmetical mean of the PLIs of the comparables which were accepted by it. In this case such comparables happen to be those of the respondent/assessee. The Tribunal, in selecting only one PLI out of a set of PLIs had clearly erred in law. However, in the facts of the present case that would not make any difference to the respondent/assessee's case in as much as even if the arithmetical mean of the comparables as accepted by the Tribunal is taken into account, the PLI would, whether the seven companies are taken into consideration or all eight companies are taken into consideration, be less than 6.99 per cent which is the PLI of the respondent/assessee for the relevant year, that is, financial year ending 31st March, 2002. We may also make it clear that the reference to the OECD Guidelines by the Tribunal in the impugned order are in the context of the reliance placed by the TPO on the very same guidelines, in particular, to para 3.27 thereof. In the present case, there are specific provisions of sub- rr. (2) and (3) of r. 10B of the said rules as also of the first proviso to s. 92C(2) of the said Act which apply. Therefore, the question of applying OECD Guidelines does not arise at all.‖

It is therefore, evident that the Special Bench and this Court stressed that mere distortion cannot be the basis of exclusion, given the mandate of Section 92C. The assessee had during the hearing, heavily relied on OECD guidelines and another Division Bench ruling in Agnity (supra). This court proposes to take up the latter decision first for discussion. In Agnity (supra), the revenue had questioned, inter alia, the ITAT decision to exclude the data relating to Infosys. One of the reasons was that the said company was a “giant” corporation and was involved in multifarious activities. After reproducing the comparative chart and noticing the facts, the Court reasoned as follows:

―6. Learned counsel for the Revenue has submitted that the Tribunal after recording the aforesaid table has not affirmed or given any finding on the differences. This is partly correct as the Tribunal has stated that Infosys Technologies Ltd. should be excluded from the list of comparables for the reason latter was a giant company in the area of development of software and it assumed all risks leading to higher profits, whereas the respondent-assessee was a captive unit of the parent company and assumed only a limited risk. It has also stated that Infosys Technologies Ltd. cannot be compared with the respondent- assessee as seen from the financial data etc. to the two companies mentioned earlier in the order i.e. the chart. In the grounds of appeal the Revenue has not been able to controvert or deny the data and differences mentioned in the tabulated form. The chart has not been controverted.

7. Learned counsel for the appellant Revenue during the course of hearing, drew our attention to the order passed by the TPO and it is pointed out that based upon the figures and data made available, the TPO had treated a third company as comparable when the wage and sale ratio was between 30% to 60%. By applying this filter, several companies were excluded. This is correct as it is recorded in para 3.1.2 of the order passed by the TPO. TPO, as noted above, however had taken three companies, namely, Satyam Computer Service Ltd., L&T Infotech Ltd. and Infosys Technologies as comparable to work out the mean.

8. It is a common case that Satyam Computer Services Ltd. should not be taken into consideration. The Tribunal for valid and good reasons has pointed out that Infosys Technologies Ltd. cannot be taken as a comparable in the present case. This leaves L&T Infotech Ltd. which gives us the figure of 11.11 %, which is less than the figure of 17% margin as declared by the respondent-assessee. This is the finding recorded by the Tribunal. The Tribunal in the impugned order has also observed that the assessee had furnished details of workables in respect of 23 companies and the mean of the comparables worked out to 10%, as against the margin of 17% shown by the assessee. Details of these companies are mentioned in para 5 of the impugned order.

9. In view of the aforesaid position, we do not think that any substantial question of law arises for consideration. The appeal is dismissed.‖

28. Quite evidently, the Court accepted the assessee‟s contentions with respect to dissimilarity of comparables; given the facts, equally, there was sufficient material to favour that view, in the facts of the case. The Court, unlike in Mentor Graphics (supra) did not undertake an analysis of the provisions involved- it was not also necessary, given the admitted state of facts.

29. Considerable inspiration was drawn from OECD guidelines to say that extraordinary facts in relation to a comparable should lead to its rejection in the TP analysis. The relevant provisions of the 2010 OECD Transfer Pricing Guidelines are extracted below:

―A.7.3 – Extreme Results: Comparability Considerations 3.63: Extreme results might consist of losses or unusually high profits. Extreme results can affect the financial indicators that are looked at in the chosen method (e.g. the gross margin when applying a resale price, or a net profit indicator when applying a transactional net margin method). They can also affect other items, e.g. exceptional items which are below the line but nonetheless may reflect exceptional circumstances. Where one or more of the potential comparables have extreme results, further examination would be needed to understand the reasons for such extreme results. The reason might be a defect in comparability, or exceptional conditions met by an otherwise comparable third party. An extreme result may be excluded on the basis that a previously overlooked significant comparability defect has been brought to light, not on the sole basis that the results arising from the proposed ―comparable‖ merely appear to be very different from the results observed in other proposed ―comparables‖.

3.64: An independent enterprise would not continue loss- generating activities unless it had reasonable expectations of future profits. See paragraphs 1.70 to 1.72. Simple or low risk functions in particular are not expected to generate losses for a long period of time. This does not mean however that loss- making transactions can never be comparable. In general, all relevant information should be used and there should not be any overriding rule on the inclusion or exclusion of loss-making comparables. Indeed, it is the facts and circumstances surrounding the company in question that should determine its status as a comparable, not its financial result.

3.65: Generally speaking, a loss-making uncontrolled transaction should trigger further investigation in order to establish whether or not it can be a comparable. Circumstances in which loss-making transactions/enterprises should be excluded from the list of comparables include cases where losses do not reflect normal business conditions, and where the losses incurred by third parties reflect a level of risks that is not comparable to the one assumed by the taxpayer in controlled transactions. Loss-making comparables that satisfy the comparability analysis should not however be rejected on the sole basis that they suffer losses.

3.66: A similar investigation should be undertaken for potential comparables returning abnormally large profits relative to other potential comparables.‖

On the use of multiple year data, this is what the said guidelines provide: ―B.5 Multiple Year Data

3.75: In practice, examining multiple year data is often useful in a comparability analysis, but it is not a systematic requirement. Multiple year data should be used where they add value to the transfer pricing analysis. It would not be appropriate to set prescriptive guidance as to the number of years to be covered by multiple year analyses.

3.76: In order to obtain a complete understanding of the facts and circumstances surrounding the uncontrolled transaction, it generally might be useful to examine data from both the year under examination and prior years. The analysis of such information might disclose facts that may have influenced (or should have influenced) the determination of the transfer price. For example, the use of the data from past years will show whether a taxpayer‘s reported loss on a transaction is part of a history of losses on similar transactions, the result of particular economic conditions in a prior year that increased costs in the subsequent year, or a reflection of the fact that a product is at the end of its life cycle. Such an analysis may be particularly useful where a transaction profit method is applied. See paragraph 1.72 on the usefulness of multiple year data in examining loss situations. Multiple year data can also improve the understanding of long term arrangements.

3.77: Multiple year data will also be useful in providing information about the relevant business and product life cycles of the comparables. Differences in business or product life cycles may have a material effect on transfer pricing conditions that needs to be assessed in determining comparability. The data from earlier years may show whether the independent enterprise engaged in a comparable transaction was affected by comparable economic conditions, or whether different conditions in an earlier year materially affected its price or profit so that it should not be used as a comparable.

3.78: Multiple year data can also improve the process of selecting third party comparables, e.g. by identifying results that may indicate a significant variance from the underlying comparability characteristics of the controlled transaction being reviewed, in some cases leading to the rejection of the comparable, or to detect anomalies in third party information.

3.79: The use of multiple year data does not necessarily imply the use of multiple year averages. Multiple year data and averages can however be used in some circumstances to improve reliability of the range. See paragraphs 3.57-3.62 for a discussion of statistical tools.‖

30. The reasoning adopted in various judgments noticed above, shows that functional analysis seeks to identify and compare the economically significant activities and responsibilities undertaken, assets used and risks assumed by the parties to the transaction. Quantitative and qualitative filters/criteria have been used in different cases to include or exclude comparables. The intuitive logic for excluding big companies from the list of comparables while undertaking the FAR analysis of a smaller company is attractive, given that such big companies provide services to diverse clientele, perform multifarious functions, often assume risks and employ intangible assets which are specially designed, unlike in the case of smaller companies. The bigger companies have an established reputation in the segment, are well known and employ economies of scale to a telling end. On the other hand, these obvious – and apparent features should not blind the TPO from the obligation to carry out the transfer pricing exercise within the strict mandate of Section 92 C and Rules 10-A to 10-E.

31. Arm's length price determination, in respect of an international transaction has necessarily to confirm to the mandate of Rule 10B. In this case, the method followed for determining the arm's length price of the international transaction adopted by the assessee and the revenue is the TNMM. The comparability of an international transaction with an uncontrolled transaction has, in such cases, to be seen with reference to the functions performed, taking into account the assets employed or to be employed and the risks assumed by the respective parties to the transaction as per rule 10B(2)(b). The specific characteristics of the property transferred or services provided (contemplated by Rule 10B(2)(a)) in either transactions may be secondary, for judging comparability of an international transaction in the TNMM, because the price charged or paid for property transferred or services provided and the direct and indirect cost of production incurred by the enterprise in respect of property transferred or services provided go into reckoning comparability analysis in the transaction methods, i.e the comparable uncontrolled price, resale price and cost plus whereas the profit based method such as transactional net margin method takes into account, the net margin realised. In TNMM, comparability of an international transaction with an uncontrolled transaction is to be seen with reference to functions performed as provided in sub-rule (2)(b) of rule 10B read with sub-rule (1)(e) of that rule after taking into account assets employed or to be employed and the risks assumed by the respective parties to the transaction.

As noticed earlier, Rule 10B(3) mandates that a given or select uncontrolled transaction selected in terms of Rule 10B(2) “shall be comparable to an international transaction” if none of the differences, if any, between the compared transactions, or between enterprises entering into such transactions ―are likely to materially affect the price or cost charged or paid or the profit arising from such transaction in the open market or reasonably accurate adjustment can be made to eliminate the effects of such difference.‖

32. Now, the sequitur of Rule 10B (2) and (3) is that if the comparable entity or entity‟s transactions broadly conform to the assessee‟s functioning, it has to enter into the matrix and be appropriately considered. The crucial expression giving insight into what was intended by the provision can be seen by the use of the expression: ―none of the differences, if any, between the transactions being compared, or between the enterprises entering into such transactions are likely to materially affect the price or cost charged or paid in, .. such transactions in the open market‖. The other exercise which the TPO has to necessarily perform is that if there are some differences, an attempt to “adjust” them to “eliminate the material effects” should be made:

―(ii) reasonably accurate adjustments can be made to eliminate the material effects of such differences.‖

33. Such being the case, it is clear that exclusion of some companies whose functions are broadly similar and whose profile – in respect of the activity in question can be viewed independently from other activities- cannot be subject to a per se standard of loss making company or an “abnormal” profit making concern or huge or “mega” turnover company. As explained earlier, Rule 10B (2) guides the six methods outlined in clauses (a) to (f) of Rule 10B(1), while judging comparability. Rule 10B (3) on the other hand, indicates the approach to be adopted where differences and dissimilarities are apparent. Therefore, the mere circumstance of a company - otherwise conforming to the stipulations in Rule 10B (2) in all details, presenting a peculiar feature - such as a huge profit or a huge turnover, ipso facto does not lead to its exclusion. The TPO, first, has to be satisfied that such differences do not ―materially affect the price...or cost‖; secondly, an attempt to make reasonable adjustment to eliminate the material effect of such differences has to be made.

34. The Court is also aware of the factors mentioned in Rule 10B (2), i.e characteristics of the service provided, functions performed taking into account assets employed or to be employed and the risks assumed, by the respective parties to the transactions; contractual terms of the transactions indicating how the responsibilities, risks and benefits are to be divided between the respective parties to the transactions; conditions prevailing in the markets in which the respective parties to the transactions operate, including the geographical location and size of the markets, the laws and the Government orders in force; costs of labour and capital in the markets, overall economic development and level of competition and whether the markets are wholesale or retail. These elements comprehend the similarities and dissimilarities; clause (f) of Rule 10C(2) specifically provides that ―the extent to which reliable and accurate adjustments can be made to account for differences, if any, between the international transaction or the specified domestic transaction and the comparable uncontrolled transaction or between the enterprises entering into such transactions and the nature, extent and reliability of assumptions required to be made in application of a method‖ have to be taken into consideration by the TPO.

35. As regards the relevance of multiple year data for transfer pricing determination, this Court is of the opinion that the general rule as prescribed in Rule 10B(4) mandates the tax authorities to take into account only the relevant assessment year‟s data. The proviso to Rule 10B(4) permits data relating to two years prior to the relevant assessment year to be taken into account in the event that they have an influence on the determination of price. However, in such instances, the onus lies upon the assessee to establish the relevance of such data. The language of Rule 10B(4) does not leave any scope for ambiguity on this issue. This Court notices that this very ground- i.e applicability of previous years‟ data for reaching out comparables, was sought to be urged in Marubeni India (P) Ltd v DIT 354 ITR 638 but deliberately left moot, because the assessee had given it up before the Tribunal. The TPO in his order dated 03.10.2011 has comprehensively examined the authorities on this issue and rightly held that ordinarily, the revenue has to consider only the relevant assessment year‟s data under Rule 10B(4) and that data from earlier period may also be considered if “it reveals certain facts which have an influence on the determination of transfer prices in relation to the transaction being considered”. The assessee has placed significant reliance on the OECD guidelines to contend the admissibility of previous year‟s data for transfer pricing determination. However, for reasons given in the paragraphs below, this Court is of the opinion that the OECD guidelines have no bearing on this issue.

36. This Court holds that in the facts of the present case, the assessee was incorrect, both in its reliance placed upon previous years‟ data as well as the manner of such reliance. First, the assessee‟s justification for relying on such data is the volatility in the comparables‟ profit margins and the consequent inability to transact at a consistent ALP. However, this is not warranted herein. Whilst there may be a wide fluctuation in the profit margins of comparables from year-to-year, this by itself does not justify the need to take into account previous years‟ profit margins. The transfer pricing mechanism provided in the Act and the Rules prescribes that while determining the ALP, the arithmetic mean of all comparables is to be adopted. This is to offset the consequence of any extreme margins that comparables may have and arrive at a balanced price. Similarly, the wide fluctuations in profit margins of the same entity on a year-to-year basis would be offset by taking the arithmetic mean of all comparables for the assessment year in question. In any case, in the event that the volatility is on account of a materially different aspect incapable of being accounted for, the analysis under would Rule 10B(3) would exclude such an entity from being considered as a comparable. Secondly, as regards the manner of using previous years‟ data, the assessee has taken the arithmetic mean of the comparables‟ profit margins for the assessment year in question and two previous years. This Court disagrees. The proviso to Rule 10B(4), read with the sub-rule, itself indicates that the purpose for which previous years‟ data may be considered is - analysing the comparability of an uncontrolled transaction with an international transaction. It does not prescribe that once an uncontrolled transaction has been held to be a „comparable‟, in order to obviate an apparent volatility in the data, the arithmetic mean of three years (the assessment year in question and two previous years‟) may be taken.

That would amount to assigning equal weight to the data for each of the three years, which is against the mandate of Rule 10B(4). The use of the word „shall‟ in Rule 10B(4) and, noticeably, „may‟ in the proviso, implies that the relevant assessment year‟s data is of primary consideration, as opposed to previous years‟ data.

37. The contention that OECD guidelines have to be taken into consideration requires a closer scrutiny. The Organisation for Economic Co- operation and Development (OECD) is an international economic organisation of 34 countries founded in 1961 to stimulate economic progress and world trade. India is not a member of this grouping; it has an observer status. She has, however, of late been actively co-operating with the organization. The Guidelines of OECD therefore, have only persuasive status; they do not have any legal sanction- unlike, for instance Double Taxation Avoidance Agreements which courts are duty bound to interpret and implement, in terms of municipal law, given the compulsion of provisions of the Income Tax Act. Secondly – and more importantly- the provisions of the Constitution compel a national legislation, to embody the terms of a treaty, for it to be enforceable in courts in India. This is because of Article 253 of the Constitution and the dualist tradition (of International law) followed by India, whereby treaties by themselves are legally unenforceable in courts, but are to be assimilated through municipal (or national) legislation. Our Supreme Court has, in the area of human rights – particularly in personal liberty, been emphasizing that to the extent the provision of any treaty is in consonance with provision of the Constitution (such as Article 21) it would be read along with such provision or right (Jolly George Varghese and Anr. v. The Bank of Cochin, AIR 1980 SC 470, Apparel Export Promotion Council v. A.K. Chopra, AIR 1999 SC 625; Kubic Dariusz v Union of India AIR 1990 SC 605). Thus, the Courts are primarily bound by the law on the subject in India; if the law is clear and unambiguous, there is no question of resorting to extrinsic sources. The only rider is that if the terms of such conventions or treaties are similar to the law applicable in India, courts may consider precedents in that regard; however those are only of persuasive value.

38. The aforesaid conclusion is fortified by the Division Bench decision of this Court in Mentor Graphics (supra), where the Court noted:

―We may also make it clear that the reference to the OECD guidelines by the Tribunal in the impugned order are in the context of the reliance placed by the Transfer Pricing Officer on the very same guidelines, in particular, to paragraph 3.27 thereof. In the present case, there are specific provisions of sub- rules (2) and (3) of Rule 10B of the said Rules as also of the first proviso to section 92C(2) of the said Act which apply. Therefore, the question of applying OECD guidelines does not arise at all.‖

This Court also notes that a recent decision in Sony Ericsson Mobile Communications India Pvt. Ltd. v. CIT (dated 16.03.2015) relied extensively on the OECD Guidelines. However, the said ruling itself recognized that the provisions of the Act and the Rules “are supreme”. Therefore, this Court holds that where they (i.e., the Act and the Rules) adequately cover a field, reliance on the OECD Guidelines is not warranted. At this stage, we deem it fit to quote the following observations of the Supreme Court in Entertainment Network (India) Ltd. v. Super Cassette Industries Ltd., (2008) 13 SCC 30:

―However, applicability of the International Conventions and Covenants, as also the resolutions, etc. for the purpose of interpreting domestic statute will depend upon the acceptability of the Conventions in question. If the country is a signatory thereto subject of course to the provisions of the domestic law, the International Covenants can be utilized. Where International Conventions are framed upon undertaking a great deal of exercise upon giving an opportunity of hearing to both the parties and filtered at several levels as also upon taking into consideration the different societal conditions in different countries by laying down the minimum norm, as for example, the ILO Conventions, the court would freely avail the benefits thereof. Those Conventions to which India may not be a signatory but have been followed by way of enactment of new Parliamentary statute or amendment to the existing enactment, recourse to International Convention is permissible.‖ (emphasis supplied)

The above excerpt indicates that courts must be cautious of relying upon international conventions to which India is not a signatory and with respect to which there is no legislative mandate whatsoever.

In any event, the OECD Guidelines relevant herein are in consonance with the Rules. Para 3.63 of the Guidelines states that an extreme comparable cannot be excluded “on the sole basis that the results arising from the proposed ̳comparable‘ merely appear to be very different from the results observed in other proposed ̳comparables‘‖ and that “further examination would be needed to understand the reasons for such extreme results‖. Similarly, para 3.65 states that “loss-making comparables that satisfy the comparability analysis should not however be rejected on the sole basis that they suffer losses‖. Further, para 3.64 states that “it is the facts and circumstances surrounding the company in question that should determine its status as a comparable, not its financial result‖. The same approach is prescribed in para 3.66 for entities making supernormal profits. Therefore, both the OECD Guidelines as well as Rule 10B (2) and 10B (3) do not, in any manner, prescribe automatic exclusion of entities with extreme financial results.

Similarly, insofar as the use of multiple year data is concerned, Para 3.75 of the OECD Guidelines states that “[m]ultiple year data should be used where they add value to the transfer pricing analysis.‖ This is akin to the proviso to Rule 10B(4) which provides for “data relating to a period not being more than two years prior to such financial year [to] be considered if such data reveals facts which could have an influence on the determination of transfer prices in relation to the transactions being compared.‖ Crucially, as noted by the TPO, para 3.79 of the Guidelines states that the “use of multiple year data does not necessarily imply the use of multiple year averages”. Thus, even if multiple year data is taken into consideration while determining the arm‟s length price, it may only be for the purposes of factoring in material changes in, inter alia, economic conditions, third party variables, etc.

39. This Court proceeds on the basis that there is sufficient guidance and clarity in Rule 10B on the principles applicable for determination of ALP. These include the various factors to be taken into consideration, approach to be adopted (functions performed, taking into account risks borne and assets employed, size of the market, the nature of competition, terms of labour, employment and cost of capital, geographical location etc). The extent of accurate adjustments possible, too, is a factor to be considered. Rule 10B (3) then underlines what the ALP determining exercise entails, if there are dissimilarities which materially affect the price charged etc: the first attempt has to be to eliminate the components which so materially affect the price or cost. In other words, given the data available, if the distorting factor can be severed and the other data used, that course has to be necessarily adopted.

40. In the present case, this Court holds that once Brescon, Keynote and Khandwala Securities are held to be functionally similar to the assessee, they would be included as comparables, notwithstanding their high profit margins, provided that the material difference on account of such high profit margins can be eliminated under the Rule 10B(3) analysis.

41. This Court, on a perusal of the orders of the lower authorities and the assessee‟s submissions before them which have been placed on record in this appeal, finds that the assessee‟s contentions with respect to the exclusion of Brescon and Khandwala Securities were based only on their exceptionally high profit margins for the assessment year in question and not on the grounds of functional dissimilarities. Indeed, the assessee did not contend the latter before the lower authorities. The assessee has sought to highlight differences in the risk profiles of the assessee and Brescon in the present appeal. However, this Court holds that such a contention cannot beraised for the first time at this stage. Therefore, Brescon and Khandwala Securities are held to be functionally similar, and the matter is remitted to the DRP for the purposes of examination under Rule 10B(3) of the Rules.

In the event that the material differences arising out of the extremely high profits cannot be eliminated as per Rule 10B(3), these two entities will have to be discarded as comparables.

42. As far as Keynote is concerned, this Court notices that the assessee had challenged its inclusion as a comparable on two grounds: a) differences in the activities of Keynote and the assessee; and b) exceptionally high profit margins. The TPO rejected the first ground relying on the fact that the assessee had used it as a comparable for previous years and in the subject assessment year as well, it qualified as a comparable based on the assessee‟ssearch process. Further, the TPO held that Keynote was engaged in financial consultancy and would therefore be considered as a comparable. The ITAT, for reasons unknown, did not examine this issue. This Court notes that the assessee is engaged in the business of rendering financial research and advisory services. It is responsible for investigation and advice to some of its group companies on structuring potential investments and exit opportunities; advising the group companies of investment and disposition opportunities; collection and dissemination of financial information of prospective entities; and other related services. On the other hand, Keynote, as per its Directors‟ Report for FY 2007-08, is involved in “Lead Managing IPOs, Rights Offers, Buybacks and Takeovers. [It] also expanded its reach in Corporate Finance & M&A Advisory.” The services provided by Keynote also include managing public issue of securities, underwriting, project appraisal, equity research, capital restructuring, loan and lease syndication, placement services, portfolio management, debenture trustee, managing/advising on international offerings of debt/equity, private placement of securities, etc.

Evidently, the assessee does not provide any of these services enumerated above. Given such functional differences and the mandate of Rule 10B(2)(b), there could be merit in the argument that Keynote cannot be considered a comparable for determining the ALP. The fact that the assessee had included it in the previous assessment years does not have any bearing on its inclusion for the subject assessment year. In this regard, this Court relies on the Supreme Court‟s decision in Commissioner of Income Tax v. C. Parakh & Co (India) Ltd, 29 ITR 661, where the Court noted:

―Whether the respondent is entitled to a particular deduction or not will depend on the provision of law relating thereto, and not on the view which it might take of its rights, and consequently, if the whole of the commission is under the law liable to be deducted against the Indian profits, the respondent cannot be estopped from claiming the benefit of such deduction, by reason of the fact that it erroneously allocated a part of it towards the profits earned in Karachi. What has therefore to be determined is whether, notwithstanding the apportionment made by the respondent in the profit and loss statements, the deduction is admissible under the law.‖

Further, a Division Bench of this Court in CIT v. Bharat General Reinsurance, 81 ITR 303 has also held that there is no estoppel against law under the Act. The Court therein held as follows:

―It is true that the assessee itself had included that dividend income in its return for the year in question but there is no estoppel in the Income tax Act and the assessee having itself challenged the validity of taxing the dividend during the year of assessment in question, it must be taken that it had resiled from the position which it had wrongly taken while filing the return. Quite apart from it, it is incumbent on the income-tax department to find out whether a particular income was assessable in the particular year or not. Merely because the assessee wrongly included the income in its return for a particular year, it cannot confer jurisdiction on the department to tax that income in that year even though legally such income did not pertain to that year.‖

For the sake of completion, this Court would also deal with the assessee‟s reliance on the DRP‟s order dated 04.03.2013 (for AY 2006-07) for the exclusion of Keynote as a comparable. The DRP directed such exclusion on two grounds: a) the fact that Keynote was making exceptionally high profits; and b) only single year data could be considered for determining ALP in the present case and the volatility in profit margins of Keynote would distort the ALP. Thus, the DRP did not examine the functional comparability of Keynote with the assessee.

In light of the discussion above, this Court remits the matter for consideration to the DRP to properly apply the test indicated in this judgment and analyse the functional similarity of Keynote with the assessee. In the event that the DRP finds them to be functionally comparable, it would proceed to carry out the Rule 10B(3) analysis as in the case of Khandwala Securities and Brescon.

43. The final question that arises for this Court‟s determination in the present appeal is the assessee‟s claim for deduction under Section 36(1)(ii) of the Act in respect of the bonus paid by it to its two shareholders - Ashish Dhawan and Kunal Shroff. The lower authorities denied such claim, holding that the bonus was paid to the shareholders in lieu of dividend with the objective of avoiding tax. Such inference was drawn from two facts: a) the bonus paid was in proportion of their shareholding in the assessee company, i.e. 2:1; and b) no dividend had been declared by the assessee.

However, a perusal of an excerpt from the DRP‟s order dated 21.09.2012 quoted by the AO in his order dated 19.10.2012 contradicts both these facts: a) bonus was not paid in the ratio of 2:1 and b) the assessee had declared interim dividend of ` 5,47,47,000/-. Further, the bonuses paid to the two shareholder-directors in the preceding two financial years were in the ratio of 60-65%:40-35%, even though their shareholding was 1:1. The balance sheet of the assessee placed on record also indicates that the two shareholders also hold directorial positions in the assessee. Therefore, the assessee‟s contention that the bonus was paid to the shareholders in their managerial capacity, like in the case of other managers, cannot be questioned merely on the basis of a speculation by the revenue that such payment was to avoid tax. In such circumstances, the deduction under Section 36(1)(ii) in respect of payment of bonus to the two shareholder-directors is allowed. The assessee has relied upon a number of judicial pronouncements to support its contention.

However, we do not consider it necessary to discuss those decisions for ruling in its favour. Therefore, this question is answered in favour of the assessee.

44. In light of the above findings, this Court concludes as follows:

a. The mere fact that an entity makes high/extremely high profits/losses does not, ipso facto, lead to its exclusion from the list of comparables for the purposes of determination of ALP. In such circumstances, an enquiry under Rule 10B(3) ought to be carried out, to determine as to whether the material differences between the assessee and the said entity can be eliminated. Unless such differences cannot be eliminated, the entity should be included as a comparable.

b. While determining the comparability of transactions, multiple year data can only be included in the manner provided in Rule 10B(4). As a general rule, it is not open to the assessee to rely upon previous year‟s data.

c. As regards Khandwala Securities and Brescon, the matter is remitted to the DRP to carry out the analysis under Rule 10B(3) and determine whether the material differences arising out of their exceptionally high profits can be eliminated. If not, the said entities cannot be included as comparables. For Keynote, firstly, enquiry is to be carried out by the DRP, preceding the analysis under Rule 10B(3), as to its functional similarity with the assessee; thereafter, the exercise of determining if there are material differences on account of exceptionally high profits which are capable of elimination has to be carried out.

d. The deduction claimed by the assessee under Section 36(1)(ii) of the Act, in respect of the bonuses paid to its shareholder- employees is allowed.

45. The appeal is accordingly partly allowed, in the above terms. No costs.

S. RAVINDRA BHAT

(JUDGE)

R.K. GAUBA

(JUDGE)

APRIL 27, 2015