ORDER

K.S.S. Prasad Rao, Judicial Member. - The above two appeals are filed by the assessee having been aggrieved by the order of the learned CIT(A)-XL, Kolkata dated 12-12-2007 relating to assessment years 2003-04 and 2004-05, respectively in the case of the assessee.

2. Since the issues involved in these appeals are identical we have decided the same by this common consolidated order for the sake of convenience.

3. We first take up the ITA No. 79/Kol./2008 wherein the assessee has raised the following grounds :

"1. For that on the facts and in the circumstances, the learned CIT(A) was wrong in sustaining addition under section 43B and section 36(1)( va) of the Act on account of PF contribution towards

Aggregating to Rs. 1,29,13,290 together with arithmetical mistake in disallowance to the extent of Rs. 4,08,181 and thereby total disallowance sustained was Rs. 1,33,21,471 (Rs. 1,29,13,290 + Rs. 4,08,181).
2. For that on the facts and circumstances, the learned CIT(A) wrongly and arbitrarily disallowed payment of Rs. 67,77,210 to the approved gratuity fund.
3. (a) That on the facts and in the circumstances of the case, the learned CIT(A) erred in confirming the adjustments of Rs. 8,44,78,673 to the international transactions of the appellant with its Associated Enterprises (AE), namely Development Consultant International Ltd. (Bahamas), The Kulijan Corporation (USA) and Datacore Systems Inc. (USA), without considering the written submissions and the financials of the AE explaining the facts of the case preferred before the learned CIT(A).
(b) That on the facts and in the circumstances of the case, the learned CIT(A) erred in holding in principle that 'No matter which method is adopted, if the facts are in order, we should arrive at the same amount of adjustment', which cannot be achieved with the various prescribed methods.
4. Without prejudice to ground (3) above, in the event your Honours are of the view that the learned CIT(A) was right in confirming the adjustment even assuming but not admitting, the learned CIT(A) erred in not considering the fact that the maximum amount of adjustment to the price of international transaction of the appellant would be limited to the gross profit earned by Development Consultant International Ltd., Bahamas and accordingly should have restricted the amount of adjustment in the international transaction of the appellant with Development Consultant Ltd., Bahamas to USD 4,61,174, i.e., INR 2,07,52,830.
5. (a) That on the facts and in the circumstances of the case, the learned CIT(A) erred in specifying the reason for confirming the adjustment to the price of international transaction between the Kulijan Corporation, USA and the appellant, by way of a speaking order.
(b) Without prejudice to ground (3) above, in the event your Honours are of the view that learned CIT(A) was right in confirming the adjustment, even assuming but not admitting, the learned CIT(A) erred in not considering the fact that the maximum amount of adjustment to the price of international transaction of the appellant would be limited to the gross profit earned by the Kulijan Corporation, USA and accordingly should have restricted the adjustment to the international transaction of the appellant with the Kulijan Corporation, USA to USD 1,39,599, i.e., INR 62,81,955."

4. In ITA No. 80/Kol./2008, the assessee has raised the following grounds:

"(1) For that on the facts and in the circumstances of the case, the learned CIT(A) was wrong in sustaining addition under section 2( 24)( x) and section 36(1)( va) of the Act on account of PF contribution towards employees contribution Rs. 54,80,506.
(2) That on the facts and in the circumstances of the case, the learned CIT(A) erred in confirming the adjustments of Rs. 8,48,82,872 to the international transactions of the appellant with its associated enterprises, namely Development Consultant International Ltd. (Bahamas), The Kulijan Corporation (USA) and Datacore Systems Inc. (USA), without considering the written submissions and the financials of the associated enterprises explaining the facts of the case preferred before the learned CIT(A).
(3) Without prejudice to ground (2) above, in the event your Honours are of the view that learned CIT (A) was right in confirming the adjustment, even assuming but not admitting, the learned CIT(A) erred in not considering the fact that the maximum amount of adjustment to the price of international transaction of the appellant would be limited to the gross profit earned by Development Consultant International Ltd., Bahamas and accordingly should have restricted the amount of adjustment in the international transaction of the appellant with Development Consultant International Ltd., Bahamas to USD 1,74,086, i.e., INR 78,33,870.
(4) (a) That on the facts and in the circumstances of the case, the learned CIT(A) erred in specifying the reason for confirming the adjustment to the price of international transaction between Kulijan Corporation, USA and the appellant, by way of a speaking order.
(b) Without prejudice to ground (2) above, in the event your Honours are of the view that learned CIT(A) was right in confirming the adjustment, even assuming but not admitting, the learned CIT(A) erred in not considering the fact that the maximum amount of adjustment to the price of international transaction of the appellant would be limited to the gross profit earned by the Kulijan Corporation, USA and accordingly should have restricted the adjustment in the international transaction of the appellant with the Kulijan Corporation, USA to USD 1,28,313, i.e., INR 57,74,085."

5. At the outset of the appellate proceedings before us, the learned counsel for the assessee has submitted that he is not pressing ground No. 1(a) for assessment year 2003-04 and ground No. 1 for assessment year 2004-05, hence they are dismissed as not pressed.

6. Arguing on the ground No. 1(b), the learned counsel for the assessee has submitted that the assessee in this case has paid employees' contribution to PF within the due date of filing of the return therefore, the claim of the assessee should have been allowed in view of the decision of the Tribunal Special Bench, Kolkata in the case of Jt. CIT v. ITC Ltd. [2008] 112 ITD 57 .

7. The learned Departmental Representative for the revenue has not raised serious objection to such submission of the learned Authorised Representative for the assessee.

8. We after hearing both the parties and perusing the material available on record direct the Assessing Officer to verify the claim of the assessee and if the employees' contribution have been paid either before the filing of return or before the due date of filing of the return, whichever is earlier, then the claim of the assessee on account of employer's contribution should be allowed following the judgment of the Tribunal, Special Bench, Kolkata in the case of ITC Ltd. (supra). We also direct the Assessing Officer to verify the claim of the assessee regarding arithmetical mistake in disallowance to the extent of Rs. 4,08,181. We hold and direct accordingly and accept the ground raised by the assessee for statistical purposes.

9. Arguing on the ground No. 2, the learned counsel for the assessee has disputed the order of the learned CIT(A) in disallowing the payment of Rs. 67,77,210 to the approved gratuity fund. The learned CIT(A) in this case has observed that the assessee had claimed gratuity payment of Rs. 31,02,185. On being asked, the assessee had claimed that the deduction of gratuity of Rs. 31,02,185 has been allowed under section 143(3) and, therefore, cannot be allowed and added back to the income. The learned CIT(A), however, observed that the entire claim of gratuity in this year and for subsequent year, details of which was furnished by the assessee, which stood at Rs. 67,77,210, is not to be allowed and has directed the Assessing Officer to add back the same.

10. In appeal before us, the learned counsel for the assessee has disputed such addition and has admitted that this is a continuous practice of assessee and has claimed the gratuity payment on the basis of calculation and following the consistency of making provision and claiming the same in its return of income and, therefore, the action of learned CIT(A) in enhancing the disallowance is most arbitrary and liable to be deleted.

11. The learned Departmental Representative for the revenue has relied heavily on the order of the learned CIT(A).

12. We after hearing both the parties, are of the opinion that the claim of the assessee that the claim of gratuity has been made strictly following the consistency in the earlier years, needs verification and, therefore, we restore the matter back to the file of Assessing Officer to decide the same afresh and affording reasonable opportunity to the assessee of being heard. We hold accordingly and allow the ground for statistical purpose.

13. We now take up ground Nos. 3, 4 and 5 for the appeal relating to the assessment year 2003-04 and ground Nos. 2, 3 and 4 for the appeal relating to the assessment year 2004-05, since the said grounds deal with a common issue relating to additions made on account of transfer pricing adjustments. Incidentally, the Transfer Pricing Officer (TPO) and accordingly, the Assessing Officer, had made additions of Rs. 8,44,78,673 and Rs. 8,48,82,872 on account of transfer pricing adjustments for the assessment years 2003-04 and 2004-05, respectively. We have observed that the additions for both the years are of the same type and therefore we are of the view that all the grounds relating to the two assessment years may be disposed of together.

14. The facts are that the assessee is a company registered in India, inter alia, having three related parties in the form of subsidiaries and further step down subsidiaries incorporated in foreign countries, all of which are admittedly 'Associated Enterprises' (AE), within the meaning of section 92A of the Income-tax Act. The said AEs are as follows (a) Development Consultant International Ltd. (DCIL), a company incorporated in Bahamas; (b) The Kulijan Corporation (TKC), a company incorporated in USA; and (c) Datacore Systems Inc. (Datacore US), another company incorporated in US. The assessee had entered into the following transaction with the said AEs during the previous years relevant to the assessment years 2003-04 and 2004-05 :

Since all the above transactions constituted 'international transactions' within the meaning of section 92B(1) of the Income-tax Act, any income arising from the same in the hands of the assessee would need to be computed having regard to the Arm's Length Price (ALP) as per the provisions of section 92(1) of the Act. Incidentally, sections 92 to 92F of the Income-tax Act read with rules 10A to 10E of the Income-tax Rules, constitute a specific code or legislation on account of transfer pricing.

15. As per the provisions of section 92E of the Income-tax Act, read with section 92D of the Income-tax Act any assessee entering into any cross-border transaction with an AE is required to furnish a certificate in the form of accountant report along with the return of income and also maintain necessary documents, which must be filed with the Revenue Officer at the time of transfer pricing assessment, as and when called for. It is found that the assessee had failed to furnish the accountant's report under section 92E of the Income-tax Act, along with its return of income for the assessment year 2003-04. Further, when required by the TPO, during the course of the transfer pricing assessment for the assessment year 2003-04 to substantiate the transactions with the AEs having regard to the ALP through filing of necessary documentation, the assessee had not furnished any details/explanation with TPO. As a result of non-compliance on the part of the assessee, the TPO virtually framed a best judgment assessment so far as transfer pricing is concerned for the assessment year 2003-04 and computed a transfer pricing adjustment of Rs. 8,44,78,673 for the said assessment year. We have also noted that the assessee had again failed to furnish the accountant's report under section 92E of the Income-tax Act for the assessment year 2004-05. However, it had furnished necessary documents in the form of a report prepared by M/s. Priciwaterhouse Coopers (P.) Ltd. (PwC) at the time of the assessment proceedings for the assessment year 2004-05 before the TPO, when the TPO required the assessee to substantiate the international transactions having regard to the ALP for the said assessment year. However, the TPO had ignored the justification made by the assessee to the said report that all the transactions with the AEs were at arm's length and proceeded to inflict a similar addition worth Rs. 8,48,82,872 for assessment year 2004-05 as well, by adopting an approach similar to the one in the assessment year 2003-04. The CIT(A) had confirmed the additions made by the TPO and accordingly the Assessing Officer in the appeals relating to both the said assessment years, whereupon the assessee has filed second appeal before the Tribunal.

16. Shri Rahul Mitra and Shri Vrajesh Dutia of PwC, being Authorized Representatives (ARs) argued the case on behalf of the assessee before us by filling a detailed written submission. Shri Mitra fairly did not dispute the aforesaid facts of non-compliance on part of the assessee. However, he submitted that the assessee had filed all the necessary documents in the form of detailed documentation reports prepared by PwC, containing the analysis for computing the ALP and other relevant information at the time of hearing of the appeals relating to both the assessment years before the CIT(A) and further, the CIT(A) had forwarded all the said documents and evidences to the TPO and the Assessing Officer by way of remand proceedings, requesting for the views of the TPO and the Assessing Officer through remand reports. The learned Authorised Representative submitted that the TPO and accordingly the Assessing Officer had duly consi-dered all the said documents and evidences and furnished remand reports before the CIT(A), copies of which were filed with the paper book. However, it was a different matter altogether that the Revenue Officers had refused to accept the explanations given by the assessee and suggested to retain the initial additions made by them at the time of the regular assessments. Shri Mitra therefore submitted that the CIT(A) had duly considered the merits of the case and decided the issues against the assessee for both the assessment years, not on the ground of non-compliance but on the ground of not being satisfied with respect to the arguments of the assessee that the transactions entered into with the AEs were at arm's length. We have carefully considered the said submissions made by the learned Authorised Representative and find the same to be correct inasmuch as the CIT(A) did not uphold the additions on the ground of non-compliance by the assessee or furnishing of inadequate information/explanation by the assessee at the stage of the appellate proceedings, but the additions were sustained purely on merits. We also find that adequate opportunity was also given to the Revenue Officers at the time of the remand proceedings to verify the documents and other evidences filed by the assessee before the CIT(A) and therefore it is nobody's case that while the appeals are sub judice before the Tribunal, the additions can be sustained on the ground that the lower authorities did not have a chance to verify in depth and detail, the documents and other explanations/information filed by the assessee. We therefore, asked both the Authorised Representatives of the assessee and the learned Departmental Representative to restrict the arguments on the merits of the case.

17. Now coming to the merits of the case, Shri Rahul Mitra made the following submissions :

(1) For the assessment year 2003-04, the TPO, admittedly, being constrained due to paucity of information, had adopted a global approach by computing the assessee's net profit margin as a whole including all its domestic and third party transactions and comparing the same with the mean net profit margins of the comparable companies selected using the 'Prowess' database which is a publicly available database. Based on the aforesaid analysis, the TPO made an adjustment in ALP of the assessee and accordingly made an upward adjustment of Rs. 8,44,78,673 in the ALP of the assessee.
(2) For the assessment year 2004-05, the assessee had furnished the necessary documentation to the TPO to substantiate the international transactions having regard to the ALP for the said assessment year. The assessee had determined the ALP of each of its international transactions separately, however, the TPO ignored the separate analysis made by the assessee for each of the international transactions by the assessee and adopted a global approach whereby all the international transactions of the assessee as documented in the said report were aggregated and combined gross profit margin over direct and indirect costs for such international transactions was computed and compared with the gross margins earned by the assessee for transactions with third parties. Based on such global approach, the TPO made an adjustment in ALP of the assessee and accordingly made an upward adjustment of Rs. 8,48,82,872 in the ALP of the assessee.
(3) Shri Rahul Mitra submitted that ideally in order to arrive at the most precise approximation of fair market value, the ALP should be determined on a transaction-by-transaction basis. Only in case when the separate transactions are so closely linked or continuous that they cannot be evaluated adequately on a separate basis, that such transactions may be evaluated together. In this case, the assessee had entered into the following types of transactions ( a) Engineering drawing and design services ( b) Deputation of employees, ( c) Reimbursement of travelling costs and ( d) Rendering data entry services through its group entity Datacore India. Hence, the ALP of each of the international transactions should be determined separately as the nature of transactions entered by the assessee with its AEs was different. Hence, the ALP would be determined based on the nature of services provided by the assessee for each class of transactions taking into consideration the functions performed, assets employed and the risks assumed, by the respective parties to the transactions.
(4) The learned Departmental Representative could not controvert the above arguments placed by Shri Rahul Mitra regarding determining the ALP of each international transaction separately by giving any strong argument. We are also of the same view that the ALP should be determined on a transaction-by-transaction basis and not on an aggregate basis as done by the TPO and sustained by the CIT(A).
(5) Before proceeding to discuss the determination of ALP on a transaction-by-transaction basis, Shri Mitra submitted that as per section 92C(1) of the Income-tax Act, the ALP of an international transaction is required to be determined using any of the prescribed methods, being the most appropriate method having regard to the nature of transaction or class of transactions, however, in order to determine the most appropriate method for determining the ALP, it is first necessary to select the 'tested party'. The transfer pricing legislation in India does not provide any discussion or mention of the concept of 'tested party'. In order to understand the concept of tested party we need to refer to the transfer pricing legislations of developed countries where the principles of transfer pricing have been in use for a long time and act as a guiding force for all the developing economies. The transfer pricing guidelines issued by the US Internal revenue services under section 482 provide and discuss the concept of transfer pricing. Section 1.482-5 of the US Transfer Pricing Regulations state that 'the tested party will be the participant in the controlled transaction whose operating profit attributable to the controlled transactions can be verified using the most reliable data and requiring the fewest and most reliable adjustments, and for which reliable data regarding uncontrolled comparables can be located. Consequently, in most cases the tested party will be the least complex of the controlled taxpayers and will not own valuable intangible property or unique assets that distinguish it from potential uncontrolled comparables'. Thus, in a sense, the tested party would have lesser risk as compared to the other transacting party or the real entrepreneur. The above concept of 'tested party' has also been confirmed by the Hon'ble Delhi Tribunal in the case of Ranbaxy Laboratories Ltd. v. Addl. CIT [2008] 110 ITD 428 wherein the learned Tribunal held that, "that learned CIT failed to appreciate that under the transfer pricing mechanism 'tested party' out of two parties of a multinational involved in the transaction, the least complex party not owning intangible assets is to be taken as a tested party. The taxpayer had placed sufficient material on record to show that taxpayer was a complex party whereas other associated enterprises were less complex. ....."
(6) We agree with the view that in order to determine the most appropriate method for determining the ALP it is first necessary to select the 'tested party' and the tested party will be the least complex of the controlled taxpayer and will not own valuable intangible property or unique assets that distinguish it from potential uncontrolled comparables.
(7) Based on the above discussion on 'tested party', Shri Rahul Mitra proceeded to argue the merits of the case for the international transactions entered by the assessee with each of its AE separately.
(8) First in case of transactions entered by the assessee with its AE, DCIL in Bahamas, the learned Authorised Representative stated that DCIL, is a wholly owned subsidiary company of the assessee in Bahamas. It is engaged in providing engineering services to its clients who are engaged in turnkey projects at various industrial sites. The assessee has discussed and stated in the transfer pricing documentation report that DCIL is an entrepreneur company and since its inception it has created significant marketing intangibles. It utilizes these intangibles in order to generate work and enters into contract with third party customers. It assumes all the major risks like market, price and product risks in relation to the work. The marketing efforts undertaken by DCIL over the years and the value of intangibles created cannot be measured and co-related with the costs reflected in the books of account. It is the goodwill of the enterprise developed over the years which helps it procure contracts from third parties and does not necessarily entail a year on year expenditure to the company. DCIL downloads all the design and engineering work it generates to the assessee. Further, he stated that DCIL bears all the market price and product risks. As such it also holds commercial and marketing intangibles. Therefore, the comparability adjustments that would be required if DCIL were selected as tested party would be both substantial and unreliable. Hence, in the transfer pricing analysis undertaken by the assessee submitted before the CIT(A) and duly considered and examined by the TPO, the assessee was selected as the tested party, being a service provider and least complex of the transacting entities.
(9) Shri Rahul Mitra further submitted that during the course of appellate and remand proceedings, the CIT(A) and the TPO objected and argued against accepting DCIL as the entrepreneur company and DCPL as the service provider. They argued that DCPL was the main company in the group and DCIL at best could be considered as the marketing and client facing arm for the group. Based on such objections raised by the CIT(A), the assessee suo motu submitted its analysis before the CIT(A) selecting DCIL as the tested party, for the assessment proceedings for both the years. In the hearing before us Shri Rahul Mitra submitted that he would proceed further on his arguments based on the above analysis.
(10) Shri Rahul Mitra argued that the assessee selected DCIL as the tested party and applied resale price method for determining the ALP for the international transactions entered by the assessee with DCIL as desired by the CIT(A). In order to apply the resale price method, the taxpayer characterized DCIL as a distributor of services. The learned Authorised Representative submitted that the margins of a distributor of services should ideally be captured at net level, the net margins ( e.g., return on assets, operating income to sales, and possibly other measures of net profit) are less affected by transactional differences than is the case with price, as used in the Comparable Uncontrolled Price (CUP) method or in Resale Price Method (RPM). This has also been discussed in detail in the Transfer Pricing Guidelines issued by the OECD. Para 3.27 of these guidelines stated that, 'One strength of the transactional net margin method is that net margins ( e.g., return on assets, operating income to sales, and possibly other measures of net profit) are less affected by transactional differences than is the case with price, as used in the CUP method. The net margins also may be more tolerant to same functional differences between the controlled and uncontrolled transactions than gross profit margins. Differences in the functions performed between enterprises are often reflected in variations in operating expenses.....'. Shri Rahul Mitra also submitted that without prejudice to the above, since DCIL had incurred losses at the net level and comparing margins at the net level would involve the process of determining the veracity of operating expenses incurred by DCIL, and also since the TPO intended to apply release price method and test the gross margins of DCIL, the assessee on a conservative basis computed the gross margins earned by DCIL.

18. After due consideration of all the facts, we agree with the view that gross margins of DCIL need to be compared with gross margins of comparable uncontrolled transactions or unrelated enterprises entering into such transactions.

19. Shri Rahul Mitra then submitted that the assessee computed the margins earned by DCIL for the period from 1-4-2002 to 31-3-2003 for assessment year 2003-04 and from 1-4-2003 to 31-3-2004 for assessment year 2004-05. The results of such computation as examined by us are as follows :

20. The search process was conducted whereby comparable uncontrolled companies having similar functional profile and characteristics as DCIL were selected. This benchmarking analysis was conducted using global databases like Standard and Poor's Research Insight : Compustat North American Data, Standard & Poor's Research Insight : Compustat Global Data, Primark Disclosure's SEC, Primark Disclosure's Worldscope to identify comparable companies.

21. The entire benchmarking exercise as conducted by the assessee as filed in the paper book is extracted below :

"For assessment year 2003-04

Selection of time period :

Companies for which data for the years 2000, 2001 and 2002 were available, were selected for the purpose of our comparison. We had selected all those companies who had data for the two consecutive years mentioned above.

Keeping in view the functional profile of DCIL, the SIC codes of Major Group 50 :

Wholesale trade was selected to obtain similar comparable companies. Our search process involved distributor of services.

Selection of region :

The basic search criterias were driven by the characteristic of the entity. In order to obtain similar comparable companies in the region where DCIL operates, Canada, USA, Mexico and Peru were selected as the region for the analysis.

Basic search strategy :

Companies were extracted from the database on the basis of the abovementioned SIC codes and the region selected—this search resulted in 337 companies.

Out of the above, all companies having no financial data were excluded. This resulted in the exclusion of 40 companies and we were left with a set of 320 companies.

As mentioned earlier, the search is for distributors, whose functions performed, risks assumed and assets utilised are similar to that of DCIL, i.e., a distributor of services. Hence, in order to arrive at a set of distributors whose assets employed are more akin to that of DCIL we have applied the following quantitative filters :

Net sales 0 - The companies for which no sales data was available were rejected;
Sales, General and Administrative (SG&A) Expenses 0 - Companies having nil SG&A expenses were not considered as comparable companies to the tested party;
Research & Development Expenses/Sales 0 per cent - We have rejected companies who have incurred any research and development expenses since DCIL as a distributor would not undertake any research and developmental activity;
SG&A expenses/sales falling outside inter-quartile range - SG&A expenses/sales ratio of a distributor represents the intensity of functions performed in order to generate sales. Hence, in order to select our appropriate set of comparables we have rejected companies which were lying outside the inter-quartile range of SG&A expenses/sales of the 242 comparable companies.
Net Profit 0 - Companies having losses were rejected on a conservative basis.

On execution of the above steps a set of 98 comparable companies were left.

The above companies were then reviewed with respect to their business descriptions in order to refine the search to include in the final set of comparables, only companies with functions similar to DCIL—this resulted in the exclusion of 54 companies and we were left with a comparable set of 34 companies.

Companies were rejected by qualitative screenings due to the following reasons :

Diversified activities - Engaged in manufacturing, designing and processing activities.
Intangibles - Companies owning brand/trademarks etc.
Non-comparable business activities - Different business line or having acquired other companies.
No business description available.

The results of the database searches employing these criteria is summarized below :

Determination of arm's length results :

The arithmetical mean of the gross margins (GP/sales) of the comparable companies is 28.22 per cent.

For assessment year 2004-05.

Selection of time period :

Companies for which data for the years 2001, 2002 and 2003 were available, were selected for the purpose of our comparison. We had selected all those companies who had data for the two consecutive years mentioned above.

Basic search strategy :

Out of the above, all companies having no financial data were excluded-this resulted in the exclusion of 40 companies and were left with a set of 295 companies.

As mentioned earlier, the search is for distributors, whose functions performed, risks assumed and assets are similar to that of DCIL, i.e., a distributor of services. Hence in order to arrive at a set of distributors whose assets employed are more akin to that of DCIL we have applied the following quantitative filters :

Net Sales 0 - The companies for which no sales data was available were rejected;
Sales General and Administrative (SG&A) Expenses 0 - Companies having nil SG&A expenses were not considered as comparable companies to the tested party;
Research and Development Expenses/Sales 0 per cent - We have rejected companies who have incurred any research and development expenses since DCIL as a distributor would not undertake any research and development activity;
SG&A expenses/sales falling outside inter-quartile range - SG&A expenses/sales ratio of a distributor represents the intensity of functions performed in order to generate sales. Hence, in order to select out appropriate set of comparables we have rejected companies which were lying outside the inter-quartile range of SG&A expenses/sales of 221 comparable companies.
Net Profit 0 - Companies having losses were rejected on a conservative basis.
On execution of the above steps a set of 83 comparable companies were left. The above companies were then reviewed with respect to their business description in order to refine the search to include in the final set of comparables, only companies with functions similar to DCIL—this resulted in the exclusion of 54 companies and we were left with a comparable set of 29 companies.
Diversified activities - Engaged in manufacturing, designing and processing activities.
Intangibles - Companies owning brand/trademarks etc.
Non-comparable business activities - Different business line or having acquired other companies.
No business description available.

The results of the database searches employing these criteria are summarized below :

Determination of arm's length results :

The arithmetical mean of the gross margins (GP/sales) of the comparable companies is 28.57 per cent."

Based on the benchmarking analysis the average profitability at the gross level of such comparable companies was computed to be 28.22 per cent for assessment year 2003-04 and 28.57 per cent for assessment year 2004-05. He further submitted that the TPO and accordingly the Assessing Officer had duly considered all the said documents and evidences and furnished remand reports before the CIT(A), copies of which were filed with the paper book. He further submitted that neither the CIT(A) nor the TPO remarked against or made any objection against the benchmarking analysis performed by the assessee. However, it was a different matter altogether that the Revenue Officers suggested to retain the initial additions made by them at the time of the regular assessments.

22. The learned Authorised Representative submitted that in case of assessment year 2003-04, the arm's length GP/sales of the comparable is 28.22 per cent which is lower than the GP/sales of DCIL for the year ended 31-3-2003 of 61.10 per cent, indicating that DCIL has retained more than the arm's length margin at the gross level. He stated that transfer pricing legislation as provided in the Income-tax Act allows a taxpayer to have an option to compute the ALP which may vary from the arithmetic mean by an amount not exceeding five per cent of such arithmetic mean. For this he quoted proviso to section 92C(2) of the Income-tax Act which states that 'where more than one price is determined by the most appropriate method, the ALP shall be taken to be the arithmetical mean of such prices or, at the option of the assessee, a price which may vary from the arithmetical mean by an amount not exceeding five per cent of such arithmetical mean'. The assessee computed the ALP considering the 5 per cent tolerance range. The results of such computation are given below :

Shri Rahul Mitra submitted that based on the above analysis, the amount of adjustment in the international transaction of the assessee with DCIL would be USD 275,632. In case of assessment year 2004-05, the arm's length GP/sales of DCIL for the year ended 31-3-2004 of 25.26 per cent, indicating that DCIL has not retained more than the arm's length margin.

23. The learned Departmental Representative could not refute the arguments and facts discussed by Shri Rahul Mitra. He only reiterated that the CIT(A) during the course of appellate proceedings had argued the analysis that DCIL was the main company of substance. Based on his findings from the internet, the CIT(A) had understood the assessee to be the main company of group and DCIL, he argued was a paper company created only to evade taxes. Hence, in his order he stated that there is no justification for allowing 28 per cent margins to DCIL and entire profits should come to DCPL.

24. Shri Rahul Mitra refuted this argument of CIT(A) that DCIL was a paper company and was of no substance. He argued that the assessee had established DCIL after taking permission from the RBI and had complied with all the guidelines issued by them. He further submitted that DCIL has had business operations over so many years and was having long standing relationship with various banks. The assessee has filed copies of documents in the paper book which have been issued by the RBI showing the approval received by the assessee to invest and establish DCIL. There are also letters issued by Bank of India and Indian Overseas Bank stating the credit issued by the bank and the long standing banking relationship DCIL enjoys with these banks. Shri Rahul Mitra argued that DCIL is a company of substance and not a paper company, since no paper company would be having such substantial business operations and having excellent customer relationship with its bankers and no nationalised banks would give such substantial credit limits to a paper company.

25. We have examined all the facts and record and after due consideration of arguments placed by both the sides, we can see that DCIL has long standing banking relationship with banks like, Bank of India and Indian Overseas Bank. Further, it enjoys substantial credit limits provided by these banks for the purposes of its business operations. No paper company would be having such substantial business operations and have excellent customer relationship with its bankers and no nationalised banks would give such credit limits to a paper company. DCIL has been established after proper permission has been obtained from the RBI. Hence, based on the facts and our findings we can conclude that DCIL is a company of substance and is performing full-fledged distribution activities. It is not a paper company established to evade taxes as argued by the CIT(A) in his order. Hence, we do not find any justification in the arguments of the CIT(A) that entire profits should come to the assessee and DCIL should not retain profits. Further the benchmarking exercise and analysis conducted by the assessee has been examined by the CIT(A) and TPO but they have not been able to controvert the analysis of the assessee. Therefore, we conclude, DCIL should retain the gross margins as determined through the benchmarking exercise by the assessee discussed earlier in this order. Therefore based on such analysis, in the case of assessment year 2003-04, the amount of adjustment in the ALP of the international transactions entered by the assessee with DCIL should be restricted to USD 275,632 as submitted and conceded by the learned Authorised Representative. This amount converted to Indian rupees based on the average currency conversion rates for the relevant year comes to Rs. 1,33,42,185. In case of assessment year 2004-05, based on the analysis discussed above the assessee has conducted its international transactions with DCIL at arm's length. Hence there should not be any adjustment in the ALP for this year.

26. In case of merits of the case for International transactions by the assessee with TKC, Shri Rahul Mitra made following submissions :

(1) Similar to determining the ALP of international transactions in case of DCIL, the TPO in case of assessment year 2003-04 being constrained due to paucity of information, had adopted a global approach by computing the assessee's net profit margin as a whole including all its domestic and third party transactions and comparing the same with the mean net profit margins of the comparable companies selected using the 'Prowess' database which is a publicly available database. Based on the aforesaid analysis, the TPO made an adjustment in ALP of the assessee and accordingly made an upward adjustment of Rs. 8,44,78,673 in the ALP of the assessee.
(2) Again for the assessment year 2004-05, the assessee had furnished the necessary documentation to the TPO to substantiate the international transactions having regard to the ALP for the assessment year. The assessee had determined the ALP of each of its international transactions separately, however the TPO ignored the separate analysis made by the assessee for each of the international transaction entered by the assessee and adopted a global approach whereby all the international transactions of the assessee as documented in the said report were aggregated and combined gross profits margin over direct and indirect costs for such international transactions was computed and compared with the gross margin earned by the assessee for transactions with third parties. Based on such global approach, the TPO made an adjustment in ALP of the assessee and accordingly made an upward adjustment of Rs. 8,48,82,472 in the ALP to the assessee.
(3) The learned Authorised Representative submitted that TKC is one of the subsidiaries of AmDC Inc., New York, USA, which in turn is a subsidiary of DCIL. TKC is a multi-disciplinary engineering/architect firm. It is in consulting engineering business for last 50 years. The range of services provided by TKC includes, feasibility study, master planning, preliminary design, design development, detailed engineering, preparation of construction specs and construction drawings, bid document preparation and bid evaluation, vendors' drawings review, inspection/quality control, project management. TKC has strength of 105 technical personnel in its Philadelphia office and another 68 in its Saudi Arabia, Jordan, Egypt and other offices. It is engaged in providing engineering services to its clients who are engaged in turnkey projects at various industrial sites for customers most of which are in Middle East. Its main area of expertise is energy and power plant engineering. TKC has built up substantial experience in power plant engineering, having been responsible for over 300 electric generating plants completed or in progress, with a total capacity exceeding 30,000 MW.
(4) TKC is the entrepreneur company and has created significant marketing intangibles over the years. It uses its marketing intangibles to generate the work and assumes all the market, price and product risks. TKC carries out the work on its own, only parts of the job are sub-contracted to the assessee for its convenience. Further, being an entrepreneur company, it is difficult to determine the profits of TKC with respect to work downloaded to India (as the revenue received for work off-shored to India cannot be separately identified). Further, the revenue generated from the services provided by the assessee would form only a small part of the entire operations. The value of engineering drawing and design services rendered by the assessee to TKC for assessment year 2003-04 was Rs. 1,58,43,923 and for assessment year 2004-05 it was Rs. 1,45,77,704. The value of service forms approximately 6 per cent to 7 per cent of the cost of sales to TKC. Hence, this Shri Rahul Mitra argued, shows that testing the margins of TKC would not serve the purpose of determining the arm's length nature of the transactions undertaken by the assessee with TKC. Hence, the recourse available to test the ALP of the services rendered by the assessee to TKC is to test the margins from the Indian side. In view of the discussion on 'tested party' earlier, the assessee was selected as the tested party being least complex of the two entities. Hence, the transfer pricing analysis in this case was done from Indian side, wherein, the margins of the assessee with respect to services provided to TKC were compared internally with services provided to other third parties in foreign market.
(5) The learned Departmental Representative could not refute the submission made by Shri Rahul Mitra regarding determination of ALP of the international transactions entered by assessee with TKC by giving any strong argument. Hence, based on the facts, we see that even the CIT(A) in his order only referred to the status to TKC and did not controvert the analysis done by the assessee. Shri Rahul Mitra submitted that the TPO and accordingly the Assessing Officer had duly considered all the said documents and evidences and furnished remand reports before the CIT(A), copies of which were filed with the paper book, however, the Revenue Officers had refused to accept the explanations given by the assessee and suggested to retain the initial additions made by them at the time of the regular assessments. Based on the above facts and our findings, we agree with the arguments of Shri Rahul Mitra that in case of transactions of the assessee with TKC, the transfer pricing analysis should be done from the Indian side considering the assessee as the tested party.
(6) Shri Rahul Mitra submitted that in case of assessment year 2003-04, the analysis from the Indian side showed that the arm's length GP/DICOP of comparable uncontrolled transactions entered by the assessee with third parties is 334.17 per cent whereas the assessee has earned profitability at the gross level of 1092.45 per cent on its international transactions with TKC. In case of assessment year 2004-05, the analysis showed that the arm's length GP/DICOP of comparable uncontrolled transactions entered by the assessee with third parties is 478.75 per cent whereas the assessee has earned profitability at the gross level of 1006.67 per cent on its international transactions with TKC. This, Shri Rahul Mitra submitted, clearly established that the international transaction of the assessee in respect of engineering drawing and design services with TKC are at arm's length for both the assessment years.

27. In addition to the above transactions, the assessee had entered into a transaction in nature of deputation of employees to DCIL and TKC. Shri Rahul Mitra made the following submissions :

(1) During the years under consideration the assessee also sent its employees to TKC on deputation for a fee. The assessee did not undertake similar kind of transaction with other third parties. Hence internal cost plus method applied in case of engineering drawing and design services was not possible to be applied for this transaction. Therefore, in order to substantiate the arm's length nature of this transaction, an external benchmarking exercise was undertaken by the assessee using 'Prowess' and 'Capitaline' databases. The results of the benchmarking analysis provided in detail in the transfer pricing report submitted by the assessee with the Revenue Officers during assessment proceedings, established that the international transaction of deputation of employees to TKC was undertaken at arm's length. The results of this benchmarking exercise duly filed in the paper book showed that the assessee had earned an operating margin over total costs of 244.29 per cent in case of transactions with DCIL and 66.50 per cent in case of transactions with TKC for assessment year 2003-04. The arm's length margin for assessment year 2003-04 is 22.09 per cent. The operating margin over total costs is 173.19 per cent in case of transactions with TKC for assessment year 2004-05. The arm's length margin for assessment year 2004-05 is 19.74 per cent. This, Shri Rahul Mitra submitted, clearly established that the international transaction of the assessee in respect of deputation of employees with DCIL are at arm's length for both the assessment years.
(2) The entire benchmarking exercise as conducted by the assessee is filed in the paper book and is extracted below :

"For assessment year 2003-04

Selection of time period :

This analysis has generally focused to operating results of service providers over the two year period 2001 through 2003.

Basic search strategy :

The cut-off date of Prowess is on 15-2-2003.

Basic search criteria :

The first set in our search process was to identify a large set of potentially comparable companies based on 'company's main activity' as identified by Prowess. To ensure that we capture all the potential comparables vis-a-vis companies proving service activity, the ratio of other operating income/sales was taken to be more than 50 per cent for both the years.

The completeness of our selection was verified by checking that well- known Indian services companies were included. We also cross-checked the selected companies against those listed in the ERS database for similar industries.

Quantitative screening :

Quantitative screening is a process under which comparability is assessed by comparing economically significant financial data or ratios.

To insure comparability, companies with the following characteristics were eliminated :

(i) had a ratio of research and development (R & D) expenses to sales that exceeded 3 per cent indicating the possible ownership of intangibles and/or significant activities not involved in pure service provision;
(ii) had Net Fixed Assets (NFA) to sales that exceeded 200 per cent indicating that these companies did a significant amount of manufacturing or otherwise held substantial assets not used in providing services.
(iii) had adverse sales of less than INR 1 crore during the time period because low sales volumes may indicate that the companies are starting up operations. Moreover, the reliability of the financial data for companies with low levels of sales can be significantly reduced because the financial data for companies with low levels of sales can be significantly reduced because the same persons are often both major shareholder between profit and salaries.

Selection Process

The selected set of 375 companies was classified under various service category as per PwC definitions. About 41 board service categories were classified which are mainly Advertising and Business Promotion Services, Back-Office Services, Basic Telephony Services, Contract Manufacturing (Toll Manufacturing)—Auto parts, Contract Manufacturing (Toll Manufacturing)—Beverages, Contract Manufacturing (Toll Manufacturing) Chemicals, Contract Manufacturing (Toll Manufacturing)—Drugs, Contract Manufacturing (Toll Manufacturing) Electronics Contract Manufacturing (Toll Manufacturing)—Fabrics, Contract Manufacturing (Toll Manufacturing)—Oil, Contract Manufacturing (Toll Manufacturing) Paints, Contract Manufacturing (Toll Manufacturing) Photographic films, Contract Manufacturing (Toll Manufacturing) Tyre and rubber products, Contract Manufacturing (Toll Manufacturing)—Tools, Cellular Mobile services, Consultancy services. Software Development services—Client server and internet software solutions, Software Development services—Design and engineering software (CAD/CAE), engineering software etc.) Software Development services—Embedded systems and communication software. Software Development services—General, engineering and construction services (Dams, bridges, railways, etc.), Education services, entertainment and media services. Facilities management services, Hotel and restaurant services, Investment and financial services, industrial turnkey project execution services, internet service provider and e-commerce, Logistics support services, Media and publishing services, Medical and health care services, Oil and commercial construction services, Road transport, Shipping and cargo services, Software products and services, sales and support services, Travel related services, Warehousing and storage services.

Out of the above 41 service categories the following category was chosen to find comparable companies for DCPL :

CNS—Consultancy services after selecting this service category, the 368 companies, which did not fall within these services categories were rejected. This is shown in the table below :

The above selection process resulted in the balance 7 companies being a refined set of comparable companies. Subsequently all loss-making companies were rejected. This led to the exclusion of 2 companies leaving us with a final set of comparable of 5 companies. This set of 5 companies represented the final set of comparable companies for DCPL.

Rejection of controlled companies :

The selected set of the companies (showing a mean-OP/TC of 22.09 per cent) was tested to determine whether any controlled companies, i.e., which have foreign holding or subsidiary companies, existed. On the basis of their holding structure, three companies were determined to have controlled transactions. Since the PLI of DCPL is any way more than arm's length PLI determined after rejecting controlled companies, the controlled companies were not rejected for the purpose of this anlaysis.

The final results are tabulated bellow :—

The results of the economic analysis show that the arm's length OP/TC of the comparable companies is 22.09 per cent whereas the results of DCPL for the year ended 31-3-2003 indicate that the company has earned a profitability of 244.29 per cent with DCIL and 66.50 per cent with Kuljian on its international transactions when measured by the OP/TC criteria. This clearly establishes that the international transactions of DCPL with its associate is at arms length.

For assessment year 2004-05.

Selection of time period

This analysis has generally focused to operating results of service providers over the two-year period 2001 through 2002.

Basic search strategy :

The cut-off date of Prowess is on 16-2-2004.

We started with a universe of 7,957 companies available in Prowess updated as of 16-2-2004. From these we extracted companies, which had sales 0 in at least two out of the three financial years ending during the period 1-4-2001 and 16-2-2004. This left us with 3,560 companies. The next step in our search process was to identify a large set of potentially comparable companies in the services domain from Prowess. As the analysis pertain to capturing potential comparables within the service industry, the ratio of [oth op inc (other operating income)/sales-internal transfer-exp. sales-tax)] 50 per cent was used as a cut-off criterion for all the years. The other operating incomes definition as per Prowess captures all incomes generated by a company apart from its manufacturing and trading revenues. This left us with 582 companies within the services industry.

We also used the Capitaline Plus database to extract companies within the services industry. Here, we started with a universe of 10,954 companies available in Capitaline Plus updated as of 16-2-2004. From these we extracted companies, which had sales 0 in at least two out of the three financial years ending during the period 1-4-2001 and 16-2-2004. This left us with 3,989 companies. As, we already had a set of companies from Prowess we extracted only the extra companies from Capitaline Plus data for which were not available in Prowess. This gave us a set of 1,228 companies from Capitaline Plus. Various data fields in Capitaline Plus and Prowess were selected and the ones that were relevant to our analysis were extracted for companies, which satisfied the aforesaid basis search criteria. A simple average of the raw numeric data fields was then computed, and various computations were performed thereon as well, as on the raw data for individual years.

Quantitative screening :

Quantitative screening is a process under which comparability is assessed by comparing economically significant financial data or ratios. To insure comparability, companies with the following characteristics were eliminated:—

(i) had a ratio of Research and Development (R & D) expenses to sales that exceeded 3 per cent indicating the possible ownership of intangibles and/or significant activities not involved in pure service provision;
(ii) had Net Fixed Assets (NFA) to sales that exceeded 200 per cent indicating that these companies did a significant amount of manufacturing or otherwise held substantial assets not used in providing services;
(iii) had average sales of less than INR 1 crore during the time period because low sales volumes may indicate that the companies are starting up operations. Moreover, the reliability of the financial data for companies with low levels of sales can be significantly reduced because the same persons are often both major shareholders and key employees, diminishing the economic distinction between profits and salaries,
(iv) Qualitative selection was done on the set of companies passing the above quantitative filters, the review was based on the selecting companies providing software development services.

Table : Selection Process

1. This criterion cannot be used in Capitaline Plus because of database constraints.
2. NSEIT Ltd. got eliminated, as Prowess did not provide information on the ownership structure.

We have selected companies with functions similar to that of DCPL from the service category 'Consultancy services'. Hence, we have chosen CNS—Consultancy services as service category.

Based on the above performed search process, a comparable set of 10 companies were finally selected. The same comprised of 7 from Prowess and 3 from Capitaline Plus.

Subsequently all loss-making companies were rejected. This led to the exclusion of 3 companies leaving us with a final set of comparables of 7 companies. This set of 7 companies represented the final set of comparable companies for DCPL.

The final results are tabulated below :—

The results of the economic analysis show that the arm's length OP/TC of the comparable companies is 19.74 per cent whereas the results of DCPL for the year ended 31-3-2004 indicate that the company has earned a profitability of 173.19 per cent with DCIL and 121.09 per cent with Kuljian on its international transactions when measured by the OP/TC criteria. This clearly establishes that the international transaction of DCPL with its associate is at arms length."

28. The learned Departmental Representative could not strongly oppose the submission made by Shri Rahul Mitra. He only asked for sustaining the adjustments made by the CIT(A) in his order.

29. Based on the facts and the arguments of both the learned Authorised Representative and Departmental Representative, and after due considerations of all facts of the case, we conclude that results of the analysis done from the Indian side by the assessee show that the international transactions entered by the assessee with TKC are at arm's length. The TPO and the CIT(A) nor the TPO (sic. AO) remarked against or made any objections against the analysis performed by the assessee to justify the ALP of its international transactions entered with TKC. Therefore, the CIT(A) erred in sustaining the adjustments made by the TPO. Hence, there should not be any adjustment in respect of international transactions in nature of deputation of employees with DCIL and TKC in respect of international transactions in nature of deputation of employees with DCIL and TKC as the transaction has been entered at arm's length for both the assessment years.

30. In case of merits of the case for international transactions entered by the assessee with Datacore US, Shri Rahul Mitra made the following submissions :—

(1) Similar to the adjustments made for the international transactions with DCIL and TKC, the TPO made adjustments in the ALP of the international transactions entered by the assessee with Datacore US. The CIT(A) sustained the adjustments made by the TPO.
(2) Shri Rahul Mitra submitted that Datacore USA is the software arm of the DCPL Group. Datacore USA is the entrepreneur company engaged in providing software and back office work to its customers. It downloads part of the work relating to data entry, i.e., back office work to the assessee. The assessee in turn downloads the work to Datacore India (P.) Ltd. ('Datacore India') a 100 per cent subsidiary of the assessee. The transaction is in nature of reimbursements from the assessee's perspective and it acts merely as a pass through entity. The assessee in this case assumes no risk whatsoever and is merely a pass through entity. Datacore India, the entity actually providing the services to Datacore USA assumes normal risks of the business as service liability risk, manpower risk, and capacity utilization risk whereas Datacore USA, the entrepreneurial company assumes greater risks. With respect to reimbursements received from Datacore USA, the assessee is just a pass through entity in this case (as such no profits arise in books of the assessee), and it is Datacore India which is providing the back office services in respect of work downloaded by Datacore USA. Hence, in the books of the assessee there is no shifting of profits out of India and hence transaction has been undertaken satisfying the arm's length criteria. Hence, Shri Rahul Mitra submitted that there is no issue from transfer pricing perspective for the transaction entered by the assessee with Datacore US, it is the question of profits between two Indian entities, the assessee DCPL and Datacore India. Hence, he submitted that in order to determine whether there has been any shift in profits outside India, one needs to determine whether the margins earned by Datacore India from this transaction are at arm's length.

31. Based on the facts and our findings, we agree with the submission made by the learned Authorised Representative regarding testing the margins of Datacore India for the international transaction entered by the assessee with Datacore US.

31A. Shri Rahul Mitra submitted that, the profit margins of Datacore India were benchmarked, vide an external benchmarking exercise which has been discussed in detail in the transfer pricing study report submitted by the assessee with the CIT(A) who duly had forwarded all the said documents and evidences to the TPO and the Assessing Officer by way of remand proceedings, requesting for the views of the TPO and the Assessing Officer through remand reports. The learned Authorised Representative submitted that the TPO and accordingly the Assessing Officer had duly considered all the said documents and evidences and furnished remand reports before the CIT(A), copies of which were filed with the paper book, and neither the CIT(A) nor the TPO controverted the analysis done by the assessee and only suggested to retain the initial additions made by them at the time of the regular assessments. The results of the analysis as submitted in the records show that the arm's length OP/TC of the comparable companies is 24.08 per cent whereas the results of Datacore India for the year ended 31-3-2003 indicate that the company has earned a profitability of 26.61 per cent at the operating level on its total costs in relation to this transaction. This clearly establishes that the international transaction of Datacore India with its associate is at arm's length. In case of assessment year 2004-05, the results of the analysis show that the arm's length OP/TC of the comparable companies is 27.77 per cent whereas the results of Datacore India for the year ended 31-3-2004 indicate that the company has earned a profitability of 32.56 per cent at the operating level on its total costs in relation to this transaction. This Shri Rahul Mitra submitted, clearly established that the international transaction of the assessee with Datacore US is at arm's length.

The entire benchmarking exercise as conducted by the assessee is filed in the paper book and is extracted below :

"For assessment year 2003-04 :

The entire search procedure is the same as discussed above except for the comparable service category selected for Datacore India.

As discussed in the functional analysis section Datacore India performs functions which are purely related to back-end data entry involving no technical expertise as such. Hence we have chosen BOS—Back office services as service category. Please refer to Appendix E for the computation of the Profit Level Indicator (PLI) of Datacore India.

This resulted in the selection of six companies. Subsequently all loss-making companies were rejected. This led to the exclusion of two companies leaving us with a final set of comparables of four companies. This set of four companies represented the final set of comparable companies for Datacore India.

Adjustment to marketing and selling expenses :

A captive service provider is not required to market its services as the group companies download work to it. A third party service provider on the other hand, markets its services thereby incurring substantial expenses on account of such marketing and sales promotional activities. These third party service providers would therefore seek a return on account of the marketing functions and risks it undertakes. Therefore, services fee earned by third party service providers would have an element of profitability on account of the marketing efforts and risks undertaken by the service provider which may compose of the actual marketing expenses incurred plus an appropriate return on such expenses. This return may be considered as the Gross Profit (GP) markup on the total Value Adding Expenses (VAE) of the service provider i.e., GP/VAE. But a captive service provider does not market its services and therefore would not seek to seek profitability on account of marketing activities. The Indian databases used to conduct the comparable search generally provide information for third party service providers. Therefore, an adjustment may be required to be made to the operating results of the comparable companies to even out the differences on account of the marketing activities.

The India regulations require that the arithmetic mean of a range of comparables be used to determine the ALP of intra-group transactions. The financial results of Datacore India that the company has a OP/TC of 26.61 per cent during 2004-05, which is within the +/5 per cent range of the arithmetic mean of the 'operating profit/TC' ratio as envisaged in the proviso to section 92C(2) of the Income-tax Act, 1961. For details regarding the computation of +/5 per cent range of the arithmetic mean of comparable companies one may refer to Appendix F. This provides evidence that both the pricing basis itself of various transactions that impact the profitability of Datacore India, and the outcome of that pricing, i.e., the profitability are at arm's length and hence satisfy the 'arm's length' standard required under the Indian regulations.

Rejection of controlled companies :

The selection of four companies (showing a mean—OP/TC of 30.61 per cent) was tested to determine whether any controlled companies i.e., which have foreign holding or subsidiary company, existed. On the basis of their holding structure, three companies were determined to have controlled transactions. Since the PLI of Datacore India is any way more than the arm's length PLI determined after rejecting controlled companies, the controlled companies were not rejected for the purpose of this analysis.

The final results are tabulated below :

The results of the economic analysis show that the arm's length OP/TC of the comparable companies is 24.08 per cent after application of the +/5 per cent range whereas the results of Datacore India for the year ended 31-3-2003 indicate that the company has earned a profitability of 26.61 per cent on its international transactions when measured by the OP/TC criteria. This clearly establishes that the international transaction of Datacore India with its associate is at arm's length.

For assessment year 2004-05.

Selection of time period :

This analysis has generally focused to operating results of service providers over the two year period 2002 through 2003.

Basic search strategy :

The cut-off date of providers is 16-2-2004.

We started with a universe of 7,957 companies available in Prowess updated as of 16-2-2004. From these we extracted companies, which had sales 0 in at least two out of the three financial years during the period 1-4-2001 and 16-2-2004. This left us with 3,560 companies. The next step in our search process was to identify large set of potentially comparable companies in the services domain from Prowess. As the ratio of [oth op inc (other operating income/sales-internal transfer-exp. sales-tax)] 50 per cent was used as a cut-off criterion of all the years. The other operating income definition as per Prowess captures all incomes generated by a company apart from manufacturing and trading revenues. This left us with 582 companies within the services industry.

We also used the Capitaline Plus database to extract companies within the services industry. Here, we started with a universe of 10,954 companies available in Capitaline Plus updated as of 16-2-2004.

From these we extracted companies, which had sales 0 in at least two out of the three financial years ending during the period 1-4-2001 and 16-2-2004. This left us with 3,989 companies. As, we already had a set of companies from Prowess we extracted only the extra companies from Capitaline Plus, data for which were not available in Prowess. This gave us a set of 1,228 companies from Capitaline Plus.

Various data fields in Capitaline Plus and Prowess were selected and the ones that were relevant to our analysis were extracted companies, which satisfied basis search criteria. A simple average of the raw numeric data fields was then computed, and various computations were performed thereon as well as on the raw data for individual years.

Quantitative screening :

Quantitative screening is a process under which comparability is assessed by comparing economically significant financial data of ratios.

To insure comparability, companies with the following characteristics were eliminated :

(v) had a ratio of Research and Development (R & D) expenses to sales that exceeded 3 per cent indicating the possible ownership of intangibles and/or significant activities not involved in pure service provision;

(vi) had Net Fixed Assets (NFA) to sales that exceeded 200 per cent indicating that these companies did a significant amount of manufacturing or otherwise held substantial assets not used in providing services;

(vii) had average sales of less than INR 1 crore during the time period because low sales volume may indicate that the companies are starting up operations. Moreover, the reliability of the financial data for companies with low levels of sales can be significantly reduced because the same persons are often both major shareholders and key employees, diminishing the economic distinction between profits and salaries.

(viii) Qualitative selection was done on the set of companies passing the above quantitative filters, the review was based on the selecting companies providing software development services.

Table: Selection process

This resulted in the selection of 12 companies. Subsequently all loss-making companies were rejected. This led to the exclusion of 5 companies leaving us with a final set of comparables of 7 companies. This set of 7 companies represented the final set of comparable companies for Datacore India.

Adjustment to marketing and selling expenses :

A captive service provider is not required to market its services as the group companies download work to it. A third party service provider on the other hand, markets its services thereby incurring substantial expenses on account of such marketing and sales promotional activities. These third party service providers would therefore seek a return on account of the marketing functions and risks it undertakes. Therefore, service fee earned by third party service providers would have an element of profitability on account of the marketing efforts and risks undertaken by the service provider which may compose of the actual marketing expenses incurred plus an appropriate return on such expenses. This return may be considered as the gross profit (GP) mark-up on the total Value Adding Expenses (VAE) of the service provider i.e. GP/VAE. But a captive service provider does not market its services and therefore would not seek to keep profitability on account of marketing activities. The Indian databases used to conduct the comparable search generally provide information for third party service providers. Therefore , an adjustment may be made to the operating results of the comparable companies to even out the differences on account of the marketing activities.

The results of the economic analysis show that the arm's length OP/TC of the comparable companies is 27.77 per cent whereas the results of Datacore India for the year ended 31-3-2004 indicate that the company has earned a profitability of 32.56 per cent on its international transactions when measured by the OP/TC criteria. This clearly establishes that the international transaction of Datacore India with its associate is at arm's length."

In his order, the CIT(A) contended that the books of Datacore India showed total revenue of Rs. 7,06,73,205 for assessment year 2003-04 and Rs. 5,67,38,428 for assessment year 2004-05 whereas the appellant while testing the margins of Datacore India had only considered revenue of Rs. 1,94,91,284 for assessment year 2003-04 and Rs. 2,85,54,663 for assessment year 2004-05. Shri Rahul Mitra argued against the above contention submitting that the back office services rendered by Datacore India to Datacore US forms only a segment of the entire business operations of Datacore India. Datacore India not a captive unit of Datacore US; it also has dealings with third party customers. To substantiate this contention, Shri Rahul Mitra referred to the segmental results of Datacore India for the assessment years 2003-04 and 2004-05, which have been filed in the written submissions, as below :

Assessment Year 2003-04

Assessment Year 2004-05

The audited financials of Datacore India for the years ended 31-3-2003 and 31-3-2004 have been filed in the paper book.

32. The learned Departmental Representative could not controvert the submission made by Shri Rahul Mitra by placing any strong argument, but only reiterated that the adjustments made by the CIT(A) be sustained.

33. Based on facts and our findings of the case, after due consideration of all the facts we conclude that the analysis undertaken by the assessee to determine the ALP of the international transaction with Datacore USA is correct and on the basis of the analysis it is seen that transaction undertaken by the taxpayer with Datacore USA is at arm's length for both the assessment years.

34. Ground Nos. 4 and 5 are alternative for ground No. 3 for assessment year 2003-04 and ground Nos. 3 and 4 are alternative for ground No. 2 for assessment year 2004-05 raised by the assessee in the above two appeals. As we have found the ground No. 3 for assessment year 2003-04 and ground No. 2 for assessment year 2004-05 raised by the assessee in these appeals are meritorious, the alternative grounds relating to them were not adverted to and they are hereby dismissed as infructuous.

35. In the result, the appeals filed by the assessee are partly allowed.