|CORAM:||Rohinton Fali Nariman J., Sanjay Kishan Kaul J|
|CATCH WORDS:||Dependent Agent Permanent Establishment, Permanent Establishment, Service PE|
|DATE:||October 24, 2017 (Date of pronouncement)|
|DATE:||October 25, 2017 (Date of publication)|
|FILE:||Click here to download the file in pdf format|
|Permanent Establishment (PE) under Article 5 of DTAA: Entire law on concept of “fixed place of business”, “service PE” and “agency PE” explained. The fact that there is close association and dependence between the US company and the Indian companies is irrelevant. The functions performed, assets used and risk assumed, is not a proper and appropriate test to determine whether there is a location PE|
The assessing authority decided that the assessees had a permanent establishment (hereinafter referred to as PE) as they had a fixed place where they carried on their own business in Delhi, and that, consequently, Article 5 of the India U.S. Double Taxation Avoidance Agreement of 1990 (hereinafter referred to as DTAA) was attracted. Consequently, the assessees were liable to pay tax in respect of what they earned from the aforesaid fixed place PE in India. The CIT (Appeals) dismissed the appeals of the assessees holding that Article 5 was attracted, not only because there was a fixed place where the assessees carried on their business, but also because they were “service PEs” and “agency PEs” under Article 5. In an appeal to the ITAT, the ITAT held that the CIT (Appeals) was right in holding that a “fixed place PE” and “service PE” had been made out under Article 5, but said nothing about the “agency PE” as that was not argued by the Revenue before the ITAT. However, the ITAT, on a calculation formula different from that of the CIT (Appeals), arrived at a nil figure of income for all the relevant assessment years. The appeal of the assessees to the High Court proved successful and the High Court in e-Funds IT Solutions/ e-Funds Corp vs. DIT (Delhi High Court), by an elaborate judgment, set aside the findings of all the authorities referred to above, and further dismissed the cross-appeals of the Revenue. On appeal by the Revenue to the Supreme Court HELD dismissing the appeal:
(i) The Income Tax Act, in particular Section 90 thereof, does not speak of the concept of a PE. This is a creation only of the DTAA. By virtue of Article 7(1) of the DTAA, the business income of companies which are incorporated in the US will be taxable only in the US, unless it is found that they were PEs in India, in which event their business income, to the extent to which it is attributable to such PEs, would be taxable in India. Article 5 of the DTAA set out hereinabove provides for three distinct types of PEs with which we are concerned in the present case: fixed place of business PE under Articles 5(1) and 5(2)(a) to 5(2)(k); service PE under Article 5(2)(l) and agency PE under Article 5(4). Specific and detailed criteria are set out in the aforesaid provisions in order to fulfill the conditions of these PEs existing in India. The burden of proving the fact that a foreign assessee has a PE in India and must, therefore, suffer tax from the business generated from such PE is initially on the Revenue. With these prefatory remarks, let us analyse whether the respondents can be brought within any of the sub-clauses of Article 5.
(ii) Since the Revenue originally relied on fixed place of business PE, this will be tackled first. Under Article 5(1), a PE means a fixed place of business through which the business of an enterprise is wholly or partly carried on. What is a “fixed place of business” is no longer res integra. In Formula One World Championship Ltd. v. Commissioner of Income Tax, International Taxation-3, Delhi and others, (2017) SCC Online SC 474, this Court, after setting out Article 5 of the DTAA, held as follows:
“32. The principal test, in order to ascertain as to whether an establishment has a fixed place of business or not, is that such physically located premises have to be ‘at the disposal’ of the enterprise. For this purpose, it is not necessary that the premises are owned or even rented by the enterprise. It will be sufficient if the premises are put at the disposal of the enterprise. However, merely giving access to such a place to the enterprise for the purposes of the project would not suffice. The place would be treated as ‘at the disposal’ of the enterprise when the enterprise has right to use the said place and has control thereupon.
xxx xxx xxx
34. According to Philip Baker, the aforesaid illustrations confirm that the fixed place of business need not be owned or leased by the foreign enterprise, provided that is at the disposal of the enterprise in the sense of having some right to use the premises for the purposes of its business and not solely for the purposes of the project undertaken on behalf of the owner of the premises.
35. Interpreting the OECD Article 5 pertaining to PE, Klaus Vogel has remarked that insofar as the term ‘business’ is concerned, it is broad, vague and of little relevance for the PE definition. According to him, the crucial element is the term ‘place’. Importance of the term ‘place’ is explained by him in the following manner: “In conjunction with the attribute ‘fixed’, the requirement of a place reflects the strong link between the land and the taxing powers of the State. This territorial link serves as the basis not only for the distributive rules which are tied to the existence of PE but also for a considerable number of other distributive rules and, above all, for the assignment of a person to either Contracting State on the basis of residence (Article 1, read in conjunction with Article 4 OECD and UN MC).”
36. We would also like to extract below the definition to the expression ‘place’ by Vogel, which is as under:
“A place is a certain amount of space within the soil or on the soil. This understanding of place as a three dimensional zone rather than a single point on the earth can be derived from the French Version (‘installation fixe’) as well as the term ‘establishment’. As a rule, this zone is based on a certain area in, on, or above the surface of the earth. Rooms or technical equipment above the soil may qualify as a PE only if they are fixed on the soil. This requirement, however, stems from the term ‘fixed’ rather than the term ‘place’, given that a place (or space) does not necessarily consist of a piece of land. On the contrary, the term ‘establishment’ makes clear that it is not the soil as such which is the PE but that the PE is constituted by a tangible facility as distinct from the soil. This is particularly evident from the French version of Article 5(1) OECD MC which uses the term ‘installation’ instead of ‘place’. The term ‘place’ is used to define the term ‘establishment’. Therefore, ‘place’ includes all tangible assets used for carrying on the business, but one such tangible asset can be sufficient. The characterization of such assets under private law as real property rather than personal property (in common law countries) or immovable rather than movable property (in civil law countries) is not authoritative. It is rather the context (including, above all, the terms ‘fixed’/‘fixe’), as well as the object and purpose of Article 5 OECD and UN MC itself, in the light of which the term ‘place’ needs to be interpreted. This approach, which follows from the general rules on treaty interpretation, gives a certain leeway for including movable property in the understanding of ‘place’ and, therefore, we assume a PE once such property has been ‘fixed’ to the soil. For example, a work bench in a caravan, restaurants on permanently anchored river boats, steady oil rigs, or a transformator or generator on board a former railway wagon qualify as places (and may also be ‘fixed’). In contrast, purely intangible property cannot qualify in any case. In particular, rights such as participations in a corporation, claims, bundles of claims (like bank accounts), any other type of intangible property (patents, software, trademarks etc.) or intangible economic assets (a regular clientele or the goodwill of an enterprise) do not in themselves constitute a PE. They can only form part of PE constituted otherwise. Likewise, an internet website (being a combination of software and other electronic data) does not constitute tangible property and, therefore, does not constitute a PE. Neither does the mere incorporation of a company in a Contracting State in itself constitute a PE of the company in that State. Where a company has its seat, according to its by-laws and/or registration, in State A while the POEM is situated in State B, this company will usually be liable to tax on the basis of its worldwide income in both Contracting States under their respective domestic tax law. Under the A-B treaty, however, the company will be regarded as a resident of State B only (Article 4(3) OECD and UN MC). In the absence of both actual facilities and a dependent agent in State A, income of this company will be taxable only in State B under the 1st sentence of Article 7(1) OECD and UN MC. There is no minimum size of the piece of land. Where the qualifying business activities consist (in full or in part) of human activities by the taxpayer, his employees or representatives, the mere space needed for the physical presence of these individuals is not sufficient (if it were sufficient, Article 5(5) OECD MC and Article 5(5)(a) UN MC and the notion of agent PEs were superfluous). This can be illustrated by the example of a salesman who regularly visits a major customer to take orders, and conducts meetings in the purchasing director’s office. The OECD MC Comm. has convincingly denied the existence of a PE, based on the implicit understanding that the relevant geographical unit is not just the chair where the salesman sits, but the entire office of the customer, and the office is not at the disposal of the enterprise for which the salesman is working.”
37. Taking cue from the word ‘through’ in the Article, Vogel has also emphasised that the place of business qualifies only if the place is ‘at the disposal’ of the enterprise. According to him, the enterprise will not be able to use the place of business as an instrument for carrying on its business unless it controls the place of business to a considerable extent. He hastens to add that there are no absolute standards for the modalities and intensity of control. Rather, the standards depend on the type of business activity at issue. According to him, ‘disposal’ is the power (or a certain fraction thereof) to use the place of business directly. Some of the instances given by Vogel in this behalf, of relative standards of control, are as under:
“The degree of control depends on the type of business activity that the taxpayer carries on. It is therefore not necessary that the taxpayer is able to exclude others from entering or using the POB. The painter example in the OECD MC Comm. (no. 4.5 OECD MC Comm. on Article 5) (however questionable it might be with regard to the functional integration test) suggests that the type and extent of control need not exceed the level of what is required for the specific type of activity which is determined by the concrete business. By contrast, in the case of a selfemployed engineer who had free access to his customer’s premises to perform the services required by his contract, the Canadian Federal Court of Appeal ruled that the engineer had no control because he had access only during the customer’s regular office hours and was not entitled to carry on businesses of his own on the premises. Similarly, a Special Bench of Delhi’s Income Tax Appellate Tribunal denied the existence of a PE in the case of Ericsson. The Tribunal held that it was not sufficient that Ericsson’s employees had access to the premises of Indian mobile phone providers to deliver the hardware, software and know-how required for operating a network. By contrast, in the case of a competing enterprise, the Bench did assume an Indian PE because the employees of that enterprise (unlike Ericsson’s) had exercised other businesses of their employer. The OECD view can hardly be reconciled with the two court cases. All three examples do indeed shed some light onto the method how the relative standards for the control threshold should be designed. While the OECD MC Comm. suggests that it is sufficient to require not more than the type and extent of control necessary for the specific business activity which the taxpayer wants to exercise in the source State, the Canadian and Indian decisions advocate for stricter standards for the control threshold. The OECD MC shows a paramount tendency (though no strict rule) that PEs should be treated like subsidiaries (cf. Article 24(3) OECD and UN MC), and that facilities of a subsidiary would rarely been unusable outside the office hours of one of its customers (i.e. a third person), the view of the two courts is still more convincing. Along these lines, a POB will usually exist only where the taxpayer is free to use the POB: – at any time of his own choice; – for work relating to more than one customer; and – for his internal administrative and bureaucratic work. In all, the taxpayer will usually be regarded as controlling the POB only where he can employ it at his discretion. This does not imply that the standards of the control test should not be flexible and adaptive. Generally, the less invasive the activities are, and the more they allow a parallel use of the same POB by other persons, the lower are the requirements under the control test. There are, however, a number of traditional PEs which by their nature require an exclusive use of the POB by only one taxpayer and/or his personnel. A small workshop (cf. Article 5(2)(e) OECD and UN MC) of 10 or 12 square meters can hardly be used by more than one person. The same holds true for a room where the taxpayer runs a noisy machine.”
38. OECD commentary on Model Tax Convention mentions that a general definition of the term ‘PE’ brings out its essential characteristics, i.e. a distinct “situs”, a “fixed place of business”. This definition, therefore, contains the following conditions: – the existence of a “place of business”, i.e. a facility such as premises or, in certain instances, machinery or equipment; – this place of business must be “fixed”, i.e. it must be established at a distinct place with a certain degree of permanence; – the carrying on of the business of the enterprise through this fixed place of business. This means usually that persons who, in one way or another, are dependent on the enterprise (personnel) conduct the business of the enterprise in the State in which the fixed place is situated.”
(iii) Thus, it is clear that there must exist a fixed place of business in India, which is at the disposal of the US companies, through which they carry on their own business. There is, in fact, no specific finding in the assessment order or the appellate orders that applying the aforesaid tests, any fixed place of business has been put at the disposal of these companies. The assessing officer, CIT (Appeals) and the ITAT have essentially adopted a fundamentally erroneous approach in saying that they were contracting with a 100% subsidiary and were outsourcing business to such subsidiary, which resulted in the creation of a PE. The High Court has dealt with this aspect in some detail in which it held:
“49. The Assessing Officer, Commissioner (Appeals) and the tribunal have primarily relied upon the close association between e-Fund India and the two assessees and applied functions performed, assets used and risk assumed, criteria to determine whether or not the assessee has fixed place of business. This is not a proper and appropriate test to determine location PE. The fixed place of business PE test is different. Therefore, the fact that e-Fund India provides various services to the assessee and was dependent for its earning upon the two assessees is not the relevant test to determine and decide location PE. The allegation that e-Fund India did not bear sufficient risk is irrelevant when deciding whether location PE exists. The fact that e-Fund India was reimbursed the cost of the call centre operations plus 16% basis or the basis of margin fixation was not known, is not relevant for determining location or fixed place PE. Similarly what were the direct or indirect costs and corporate allocations in software development centre or BPO does not help or determine location PE. Assignment or sub-contract to e-Fund India is not a factor or rule which is to be applied to determine applicability of Article 5(1). Further whether or not any provisions for intangible software was made or had been supplied free of cost is not the relevant criteria/test. e-Fund India was/is a separate entity and was/is entitled to provide services to the assessees who were/are independent separate taxpayers. Indian entity i.e. subsidiary company will not become location PE under Article 5(1) merely because there is interaction or cross transactions between the Indian subsidiary and the foreign Principal under Article 5(1). Even if the foreign entities have saved and reduced their expenditure by transferring business or back office operations to the Indian subsidiary, it would not by itself create a fixed place or location PE. The manner and mode of the payment of royalty or associated transactions is not a test which can be applied to determine, whether fixed place PE exists.”
(iv) It further went on to hold that the ITAT’s finding that the assessees were a joint venture or sort of partnership with the Indian subsidiary was wholly incorrect. Also, none of these arguments have been invoked by the Revenue and such a finding would, therefore, be perverse. After citing Klaus Vogel on Double Taxation Conventions, Arvid A. Skaar in Permanent Establishment: Erosion of a Tax Treaty Principle and Bollinger vs. Commissioner, 108 S.Ct. 1173, the High Court found against the Revenue, holding that there is no fixed place PE on the facts of the present case. We agree with the findings of the High Court in this regard.
(v) Reliance placed by the Revenue on the United States Securities and Exchange Commission Form 10K Report, as has been correctly pointed out by the High Court, is also misplaced. It is clear that the report speaks of the e-Funds group of companies worldwide as a whole, which is evident not only from going through the said report, but also from the consolidated financial statements appended to the report, which show the assets of the group worldwide.
(vi) Also, Shri Ganesh has pointed out that the two American companies have four main business activities which are: ATM Management Services, Electronic Payment Management, Decision Support and Risk Management and Global Outsourcing and Professional Services.
(vii) This report would show that no part of the main business and revenue earning activity of the two American companies is carried on through a fixed business place in India which has been put at their disposal. It is clear from the above that the Indian company only renders support services which enable the assessees in turn to render services to their clients abroad. This outsourcing of work to India would not give rise to a fixed place PE and the High Court judgment is, therefore, correct on this score.
(viii) Insofar as a service PE is concerned, the requirement of Article 5(2)(l) of the DTAA is that an enterprise must furnish services “within India” through employees or other personnel. In this regard, this Court has held, in DIT v. Morgan Stanley (2007) 7 SCC 1/ 284 ITR 260, as follows:
“16. Article 5(2)(l) of DTAA applies in cases where MNE furnishes services within India and those services are furnished through its employees. In the present case we are concerned with two activities, namely, stewardship activities and the work to be performed by deputationists in India as employees of MSAS. A customer like MSCo who has worldwide operations is entitled to insist on quality control and confidentiality from the service provider. For example in the case of software PE a server stores the data which may require confidentiality. A service provider may also be required to act according to the quality control specifications imposed by its customer. It may be required to maintain confidentiality. Stewardship activities involve briefing of the MSAS staff to ensure that the output meets the requirements of MSCo. These activities include monitoring of the outsourcing operations at MSAS. The object is to protect the interest of MSCo. These stewards are not involved in day-today management or in any specific services to be undertaken by MSAS. The stewardship activity is basically to protect the interest of the customer. In the present case as held hereinabove MSAS is a service PE. It is in a sense a service provider. A customer is entitled to protect its interest both in terms of confidentiality and in terms of quality control. In such a case it cannot be said that MSCo has been rendering the services to MSAS. In our view MSCo is merely protecting its own interests in the competitive world by ensuring the quality and confidentiality of MSAS services. We do not agree with the ruling of AAR that the stewardship activity would fall under Article 5(2)(l). To this extent we find merit in the civil appeal filed by the appellant (MSCo) and accordingly its appeal to that extent stands partly allowed.
17. As regards the question of deputation, we are of the view that an employee of MSCo when deputed to MSAS does not become an employee of MSAS. A deputationist has a lien on his employment with MSCo. As long as the lien remains with MSCo the said company retains control over the deputationist’s terms and employment. The concept of a service PE finds place in the UN Convention. It is constituted if the multinational enterprise renders services through its employees in India provided the services are rendered for a specified period. In this case, it extends to two years on the request of MSAS. It is important to note that where the activities of the multinational enterprise entails it being responsible for the work of deputationists and the employees continue to be on the payroll of the multinational enterprise or they continue to have their lien on their jobs with the multinational enterprise, a service PE can emerge.
18. Applying the above tests to the facts of this case we find that on request/requisition from MSAS the applicant deputes its staff. The request comes from MSAS depending upon its requirement. Generally, occasions do arise when MSAS needs the expertise of the staff of MSCo. In such circumstances, generally, MSAS makes a request to MSCo. A deputationist under such circumstances is expected to be experienced in banking and finance. On completion of his tenure he is repatriated to his parent job. He retains his lien when he comes to India. He lends his experience to MSAS in India as an employee of MSCo as he retains his lien and in that sense there is a service PE (MSAS) under Article 5(2)(l). We find no infirmity in the ruling of ARR on this aspect. In the above situation, MSCo is rendering services through its employees to MSAS. Therefore, the Department is right in its contention that under the above situation there exists a service PE in India (MSAS). Accordingly, the civil appeal filed by the Department stands partly allowed.” (at pages 15-16)
18. It has already been seen that none of the customers of the assessees are located in India or have received any services in India. This being the case, it is clear that the very first ingredient contained in Article 5(2)(l) is not satisfied. However, the learned Attorney General, relying upon paragraph 42.31 of the OECD Commentary, has argued that services have to be furnished within India, which does not mean that they have to be furnished to customers in India. Para 42.31 of the OECD Commentary reads as under: “Whether or not the relevant services are furnished to a resident of a state does not matter: what matters is that the services are performed in the State through an individual present in that State.”
(ix) Based upon the said paragraph, Shri Venugopal has argued that in assessment year 2005-06, two employees of the American firm were seconded in India and that, therefore, it is clear that management of the American company through these employees has obviously taken place. The High Court, in dealing with this contention, has found as follows: “62. The appellants had pleaded before the authorities and the tribunal that prior to assessment year 2005-06 not even a single employee of the assessee ever visited India even for a short period and in 2005-06, two employees of e-Fund were transferred to e-Fund India and that the entire expenditure for these two employees were borne by e-Fund India. No employees were present in India after 2005-06. Presence of employees in India is relevant under Article 5(2)(l) but the said employees should furnish services within the contracting State. These services should not be mere stewardship services. The Assessing Officer has recorded that employees were seconded to e-Fund India but the functions they performed and whether they performed functions and reported to e-Fund Corp/associated enterprise was not known or ascertained. This was not the correct way of determining and deciding whether service PE existed. Whether the seconded employees were performing stewardship services or were directly involved with the working operations was relevant. It is also not known whether the services were performed related to services provided to an associated enterprise in which case clause 5(2)(l)(ii) would be applicable. In the said situation, the question of attribution of income etc. would also arise.
(x) Two employees of e-Fund Corp were deputed to e-Fund India in the assessment years 2005-06. The case of the assessee and e-Fund India is that they were deputed to look towards development of domestic work in India. Payment of these employees as per the Revenue to the extent of 25% was borne by e-Fund India and balance 75% was borne by e-Fund Corp. The Assessing Officer on this basis has observed that this reduced cost base of e-Fund India as remuneration was paid by e- Fund Corp and the said employees were at liberty to perform functions of e-Fund Corp even while working for e-Fund India. The response of the assessee as quoted in the assessment order was that e-Fund India, apart from export activities had also domestic business in India. This was evident from the return of income filed by e-Fund India where domestic income was computed separately as it was not eligible for deduction under Section 10A of the Act. Copy of the return was furnished. It was further stated that cost of personnel seconded in India was fully borne by e-Fund India i.e. 100% of the salary paid to the said employees seconded to India were debited to profit and loss accounts. 75% of the salary component was paid abroad by e-Fund Corp but the same was reimbursed by e-Fund India. This was in accordance with and permitted under the Indian Exchange Control Regulations. It was further stated that the Assessing Officer was wrong in assuming that the two seconded employees were at liberty to function for e-Fund Corp while they were working for e-Fund India. The seconded employees were working under the control and supervision of e-Fund India. The Assessing Officer thereupon has not commented on the reply of the assessee, though he has recorded comments in respect of replies to other issues raised by him (see paragraph 7 of the assessment order). The aforesaid factual assertion made by the assessee, therefore, was not negated or questioned by the Assessing Officer.”
(xi) We entirely agree with the approach of the High Court in this regard. Article 42.31 of the OECD Commentary does not mean that services need not be rendered by the foreign assessees in India. If any customer is rendered a service in India, whether resident in India or outside India, a “service PE” would be established in India. As has been noticed by us hereinabove, no customer, resident or otherwise, receives any service in India from the assessees. All its customers receive services only in locations outside India. Only auxiliary operations that facilitate such services are carried out in India. This being so, it is not necessary to advert to the other ground namely, that “other personnel” would cover personnel employed by the Indian company as well, and that the US companies through such personnel are furnishing services in India. This being the case, it is clear that as the very first part of Article 5(2)(l) is not attracted, the question of going to any other part of the said Article does not arise. It is perhaps for this reason that the assessing officer did not give any finding on this score.
(xii) Shri Ganesh has argued before us that the “agency PE” aspect of the case need not be gone into as it was given up before the ITAT. He is right in this submission as no argument on this score is found before the ITAT. However, for the sake of completeness, it is only necessary to agree with the High Court, that it has never been the case of Revenue that e-Funds India was authorized to or exercised any authority to conclude contracts on behalf of the US company, nor was any factual foundation laid to attract any of the said clauses contained in Article 5(4) of the DTAA. This aspect of the case, therefore, need not detain us any further.
(xiii) Shri Ganesh has referred to and relied upon an order of the Additional Taxation Commissioner, who is the Transfer Pricing Officer. A perusal of the above would show that a competent authority should engage in discussion with the other competent authority in a principled, fair and objective manner, with each case being decided on its own merits. It is also specifically observed that where an agreement is not otherwise achievable, then both parties should look for appropriate opportunities for compromise in order to eliminate double taxation on the facts of the case, even though a principled approach is important. The learned Attorney General also relied upon Best Practice No.1 of the said OECD Manual, which requires the publication of mutual agreements reached that may apply to a general category of taxpayers which would then improve guidance for the future. Best Practice No.1 has no application on the facts of the present case, as the agreement reached applies only to the respondent companies, and not to any general category of taxpayers. It is clear, therefore, that Shri Ganesh is right in relying upon Article 3.6 of the OECD Manual. It is very clear, therefore, that such agreement cannot be considered as a precedent for subsequent years, and the High Court’s conclusion on this aspect is also correct.
(xiv) The learned Attorney General has also laid great emphasis on non-disclosure of documents and has relied upon a long list of documents that the assessees were asked to disclose and which they did not. From this, according to the learned Attorney General, an adverse inference should be drawn, and from this alone it should be inferred that a PE of the assessees, therefore, exists in India. We are afraid that this argument cannot be countenanced at this stage as it has never been raised before any of the authorities below and has not been raised before the High Court also. This being the case, we do not think it necessary to get into this aspect of the matter.
(xv) Having held in favour of the assessees that no permanent establishment in India can possibly be said to exist on the facts of the present case, we do not deem it necessary to go into the cross-appeals that were filed before the High Court, which were dismissed by the High Court agreeing with the ITAT that the calculation of the ITAT would lead to nil taxation. This point would not arise in view of our decision on the facts of the present case. It is, therefore, unnecessary to go into this aspect of the matter.