International Taxation: The Indian Perspective [1st ed. 2020] 978-81-322-3668-9, 978-81-322-3670-2

Table of contents :
Front Matter ....Pages i-xi
Introduction (Nigam Nuggehalli)....Pages 1-3
Overview of Indian Legislation Regarding International Taxation (Nigam Nuggehalli)....Pages 5-15
Contracts, Status and the Judiciary (Nigam Nuggehalli)....Pages 17-27
Permanent Establishment (Nigam Nuggehalli)....Pages 29-46
Royalty and Fees for Technical Services (Nigam Nuggehalli)....Pages 47-63
Capital Gains (Nigam Nuggehalli)....Pages 65-73
Transfer Pricing (Nigam Nuggehalli)....Pages 75-82
General Anti-avoidance Rules (GAAR) (Nigam Nuggehalli)....Pages 83-94
BEPS, The Principal Purpose Clause and Indian Taxation (Nigam Nuggehalli)....Pages 95-107
Back Matter ....Pages 109-112

Citation preview

SPRINGER BRIEFS IN LAW

Nigam Nuggehalli

International Taxation The Indian Perspective

SpringerBriefs in Law

More information about this series at http://www.springer.com/series/10164

Nigam Nuggehalli

International Taxation The Indian Perspective

123

Nigam Nuggehalli School of Law, BML Munjal University Gurugram, India

ISSN 2192-855X ISSN 2192-8568 (electronic) SpringerBriefs in Law ISBN 978-81-322-3668-9 ISBN 978-81-322-3670-2 (eBook) https://doi.org/10.1007/978-81-322-3670-2 © The Author(s), under exclusive licence to Springer Nature India Private Limited 2020 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature India Private Limited The registered company address is: 7th Floor, Vijaya Building, 17 Barakhamba Road, New Delhi 110 001, India

Table of Cases

1. ADIT v E Funds IT Solutions Inc., (2018) 13 SCC 294. 2. Abdullabhai Abdul Kadar v. Commissioner of Income-tax Bombay City (1952 22 I.T.R. 241) 1952 Indlaw MUM 2. 3. Alembic Chemical Works Co. Ltd. v. Commissioner of Income Tax, Gujarat (1989) 177 ITR 377, 1989 Indlaw SC 559. 4. Asia Satellite Telecommunications Co. Ltd. v. DIT, 2011 SCC OnLine Del 507. 5. Assistant Commissioner of Income Tax vs. M/S Celerity Power LLP, MANU/IU/1165/2018. 6. Banca Sella S.P.A, (2016-LL-0817-75) 7. Barclays Mercantile Business Finance Limited v Mawson, [2004] UKHL 51, [2005] 1 AC 684. 8. CIT v. B.C. Srinivasa Setty, (1981) 2 SCC 460. 9. CIT v Divine Leasing and Finance, 299 ITR 268 (2007). 10. CIT v. Gautum Sarabhai Trust, 1988 SCC OnLine Guj 156. 11. CIT v. Grace Collis (Mrs.) and Others, (2001) 3 SCC 430. 12. CIT v Stellar Investments 192 ITR 287 (2007). 13. Centrica India Offshore Pvt Ltd v. CIT, (2014) 364 ITR 336. 14. Commissioner of Income Tax, Bangalore vs. B.C. Srinivasa Setty, 1981 AIR SC 972. 15. Commissioner of Income-Tax vs Davy Ashmore India Ltd., 190 ITR 626. 16. Commissioner of Income Tax, Punjab v. R.D. Aggarwal & Co., 1964 Indlaw SC 254. 17. Commissioner of Income Tax v Samsung Electronics Co Ltd., 2011 SCC OnLine Kar 3973. 18. Comptroller of Income Tax v AQQ and another appeal, [2014] SGCA 15. 19. DDIT, International Taxation v Set Satellite (Singapore) (Pte.) Ltd., MANU/IU/5227/2007. 20. DDIT v Unocol Bharat, MANU/ID/0899/2018.

v

vi

Table of Cases

21. DIC Asia Pacific Pte Ltd. v ADIT, [2012] 18 ITR (Trib) 358 (Kolkata). 22. DIT (Mumbai) v. Morgan Stanley, 2007 TII 1 SC (TP). 23. Deputy Commissioner of Income Tax (International Taxation) v. Welspun Corporation Limited, 2017 SCC OnLine ITAT 334. 24. Director of Income Tax vs. Infrasoft Ltd., 2013 SCC OnLine Del 4694. 25. Dresdner Bank v ACIT, (2006) 105 TTJ Mum 149. 26. Formula One World Championship Ltd. v CIT, International Taxation, (2017) 15 SCC 602. 27. Furniss v Dawson, [1984] A.C. 474. 28. GE India Technology Centre (P) Ltd. vs CIT [2010] 327 ITR 456 (SC). 29. HMRC v Tower M Cashback 30. Helvering v Gregory, 69 F 2nd 809 (1934). 31. Honda Siel Cars India Limited v Commissioner of Income Tax, Ghaziabad 2017 Indlaw SC 432. 32. IRC v. Duke of Westminster, [1936] AC 1 (HL). 33. Ishikawajma-Harima Heavy Industries Ltd. vs Director Of Income Tax, Mumbai, (2007) 3 SCC 481. 34. Income Tax Officer, TDS v Nokia India Pvt Ltd., 2015 SCC OnLine ITAT 9865. 35. Jindal Thermal Power Company Ltd. vs DCIT, 321 ITR 31. 36. Joseph Fowler v MNR, 1990 (2) CTC 2351. 37. Kikabai Premchand v CIT, 1953 Indlaw SC 77. 38. M/s Google India Private Ltd. vs. Addl. Commissioner of Income-tax, 2017 SCC OnLine ITAT 78. 39. Maruti Suzuki India Limited vs. CIT, 2015 SCC OnLine Del 13940. 40. Mc Dowell & Company Ltd. v The Commercial Tax Officer, [1985] 3 SCR 791. 41. McGuckian [1997] 1 WLR 991 42. M/s L.G. Electronics India Private Limited vs. Assistant Commissioner of Income Tax, MANU/ID/0036/2013. 43. Orchid Pharma Limited v Deputy Commissioner of Income Tax, Chennai, 2016 SCC OnLine ITAT 12571. 44. Raheja IT Park v CIT, 2014 11 TMI 349. 45. Ramsay v. IRC, [1982] AC 300 (HL). 46. Saroj Kumar v. I.T. Commissioner, AIR 1959 SC 1252. 47. SET Satellite v DDIT, 307 ITR 205 (2008). 48. Subhash Kabini Power v CIT, 2015 (1) TMI 646. 49. Sumitomo Mitsui Banking v DDIT, 2012 Indlaw ITAT 16. 50. Tata Consultancy Services v State of Andhra Pradesh, 271 ITR 401. 51. Union of India v. Azadi Bachao Andolan, 2003 Indlaw SC 823. 52. Union of India v Vodafone Group PLC United Kingdom and Anr, 2017 SCC OnLine Del 9930.

Table of Cases

vii

53. Vodafone International Holdings B.V. v. Union of India and Anr., [2012] 1 S. C.R. 573. 54. W.T. Ramsay Ltd. v. IRC, [1982] A.C. 300 (H.L.). 55. Wipro Ltd v DCIT, 2015 Indlaw KAR 9678.

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 3

2 Overview of Indian Legislation Regarding International Taxation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5 15

3 Contracts, Status and the Judiciary . . The English Judiciary and Tax Planning The Indian Judiciary and Tax Planning . References . . . . . . . . . . . . . . . . . . . . . .

. . . .

17 17 23 27

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

4 Permanent Establishment . . . . . . . . . . . . . . . . . . . . . . . . . Business Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ‘Business Connection’ Post-amendment . . . . . . . . . . . . . . . . Taxability of Business Profits as per DTAAs: Understanding ‘Permanent Establishments’ . . . . . . . . . . . . . . . . . . . . . . . . . Fixed PE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Service PE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dependent Agent PE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stewardship v. Deputation . . . . . . . . . . . . . . . . . . . . . . . . . . ‘Nature of Business’ and ‘Control’ Test . . . . . . . . . . . . . . . . Attribution of Profits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

........ ........ ........

29 30 31

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

33 34 34 35 36 38 41 43 46

5 Royalty and Fees for Technical Services . Introduction . . . . . . . . . . . . . . . . . . . . . . . Royalty-Definitional Provisions . . . . . . . . . Royalties and the Transfer of Software . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

47 47 48 50

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

ix

x

Contents

Retrospective Amendments to the Statutory Definition of Royalty . . . . . Fees for Technical Services-Definitional Provisions . . . . . . . . . . . . . . .

55 59

6 Capital Gains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

65

7 Transfer Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

75 82

8 General Anti-avoidance Rules (GAAR) . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The GAAR Provisions . . . . . . . . . . . . . . . . . . . . . GAAR and International Contractual Arrangements GAAR: A Comparative View . . . . . . . . . . . . . . . .

83 83 85 90 91

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

9 BEPS, The Principal Purpose Clause and Indian Taxation . BEPS: Overview and Summary . . . . . . . . . . . . . . . . . . . . . . . . The Principal Purpose Clause . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . BEPS Action Plan 6 and the PP Clause . . . . . . . . . . . . . . . . The PP Clause and the Interpretation of Tax Statutes . . . . . . Conclusion: The PPT Rule and GAAR: A Superfluous Solution to a Problem of Statutory Interpretation . . . . . . . . . . . . . . . . . . Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. 95 . 95 . 99 . 99 . 100 . 101

. . . . . . 105 . . . . . . 107

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

About the Author

Dr. Nigam Nuggehalli is Dean of the School of Law at BML Munjal University, Gurugram. Previously, he was a Visiting Professor at the National Law School, Bangalore and an Associate Professor at Azim Premji University, Bangalore. He holds a DPhil in Law from Oxford University and an LLM (Taxation) from New York University School of Law. Prof. Nuggehalli also taught at BPP University, London, United Kingdom, for several years, where he led a team of academics in delivering undergraduate programmes in jurisprudence and post graduate programmes in commercial law and international tax law. Prior to joining academia, Prof. Nuggehalli was an international tax lawyer working in New York, USA.

xi

Chapter 1

Introduction

When Sir Henry Maine said that societies progress from status to contract, he was making a point about the rules that regulate human relationships; these are largely determined by agreements rather than social rules in the modern world (Maine 1876). While Maine’s use of status was with reference to customary rules, I will assume in this paper that status can also include more complex social rules such as those that emanate from legislation and judicial decisions. Thus, while I take contracts to refer to norms arising out of enforceable agreements between individuals or entities, I take status to refer to norms arising out of customary practices, legislation and judicial decisions.1 Maine’s observation certainly applies to commercial relationships in modern societies. Contracts have pervaded the commercial lives of citizens and corporations. Contracts constitute the act of selling, leasing, servicing, incorporating, merging, acquiring, and creating anything of commercial value. Since the different parts of any sophisticated legal system are interconnected, parties’s reliance on contracts to regulate their commercial affairs has implications for the taxation of their gains and profits in business. Contracts enable parties to affect business relationships inter se in myriad ways and thus affect the tax consequences flowing from their chosen business relationships. Contracts can determine the tax unit. Taxpayers are free to enter into a Memorandum of Association (companies), a partnership agreement (partnerships and LLPs) or a Trust Deed. The unit of business can determine the applicable tax consequences. For example, a contractual relationship of shareholders would potentially result in double taxation at the corporate and the shareholder level. If parties instead contract to establish a partnership between them, there would be only one stage of taxation, at the stage of the partnership income earned by the partners. Contracts can determine the place of business activity-whether a certain activity is onshore or offshore. For example, the contract of sale can provide that the title would pass in a particular country. Contract can determine the nature of income, whether 1I

use rules and norms interchangeably in this paper.

© The Author(s), under exclusive licence to Springer Nature India Private Limited 2020 N. Nuggehalli, International Taxation, SpringerBriefs in Law, https://doi.org/10.1007/978-81-322-3670-2_1

1

2

1 Introduction

an accretion to wealth is a revenue income or capital gain, whether a given item of profit pertains to business income or some other category of income. Contracts can determine the timing of income, whether income arises in a particular financial year or the next. Contracts can have an effect on pricing of products. For example, in the case of transfer of products and services between the units of a multinational group, contracts determine the price and terms of purchase. In summary, contracts purport to determine tax consequences by establishing how people relate to each other in terms of the mode of doing business (corporate vs. non-corporate personality), wealth creation (income vs. capital), timing and pricing. Traditionally, legal systems have given priority to contracts in determining the tax consequences for taxpayers. There have been moves within legal systems to look at the overall commercial effect of all contracts that parties enter into rather than look at each contract in isolation,2 but even with such approaches, the sanctity of contracts is maintained; the only difference is that the contracts relevant to a particular controversy are looked at holistically. Therefore, until recently, contract continued to trump status, with any exceptions only going to emphasize the primacy of contracts. The Indian government, through its recent legislative interventions, has tried to reverse this primacy of contract over status. A tax system is neutral in the contract versus status debate when neither contracts nor status dominates the other, with the question of domination a question of degree rather than precise rules. Generally speaking, contracts would dominate status if, given a field of economic activity, the same economic goals can be achieved through a wide variety of substantively different contracts, with different tax consequences. Contracts are substantively different when they result in different legal consequences. For example, the same economic goal, the possession and control of industrial machinery, can be achieved through either a sale or a lease transaction, with differing tax consequences. A person who leases machinery can deduct lease payments from his business income while a purchaser of machinery cannot do so. A person who buys machinery can deduct depreciation on the machinery from his business income, while a person who leases machinery cannot do so. Generally speaking, status would dominate contracts if, given a field of economic activity, the tax law allows only one or one set of legal consequences to emerge, regardless of the kind of contractual relationships entered into by parties. For example, tax legislation might provide that if royalty is paid to a foreign person by an Indian resident, the royalty payments would be considered as Indian source income regardless of whether the contract for the payment of royalty was signed in India and regardless of whether the supplier of those technical services has a business connection with India. Either extreme in the contract versus status spectrum is problematic for the tax policy of a country. The dominance of contract over status would serve as the fulcrum of harmful tax avoidance while also diminishing any attempts by the government to regulate economic behavior through tax policy. On the other hand, the dominance of

2 W.

T. Ramsay Ltd. versus IRC, (1982) A.C. 300 (H.L.).

1 Introduction

3

status over contract would restrict economic activity and discourage entrepreneurship. With regard to economic transactions that occur across borders, the dominance of status over contract might result in a state claiming the jurisdiction to tax activities that lack genuine economic nexus with that state. The problem with the domination of status over contracts becomes acute when two strategies are used by governments to impose status over contracts. First, governments introduce deeming provisions in tax statutes to ensure that status dominates contracts through a legal fiction. Second, governments make such deeming provisions retrospective in nature, thus impacting contractual arrangements made by parties who were unaware and therefore unprepared for the deeming provisions that changed their legal position. Over the last decade, the Indian government has sought to establish the dominance of contract over status in the case of cross border economic activity. The Indian government is concerned that international tax law provides for the domination of contracts over status because of the ease with which the economic nexus rules relating to business income and capital gains can be manipulated contractually to deny a particular country tax jurisdiction over an economic activity. In order to counteract such contractual manipulation, the Indian government has introduced tax legislation that imposes a preferred legal status on the contractual transactions in cross border activity. The Indian effort to provide strength to status over contracts has the potential to result in the domination of status over contract because of the use of deeming and retrospective legislative provisions, as I will explain below. I will describe the Indian story of status versus contract through five areas of controversy: permanent establishments, FTS (Fees for Technical Services) & Royalty, capital gains, transfer pricing and general anti-avoidance rules (GAAR). The impact of the OECD BEPS (Base Erosion and Profit Shifting) Project will also be considered in this context. The area of permanent establishments demonstrates the dominance of contracts over status, at least with respect to Indian tax law. However, some recent judicial decisions in this area demonstrate the susceptibility of contract to status related arguments. The areas of FTS & Royalty as well as the areas of capital gains and transfer pricing demonstrate the Indian government’s attempt to establish, through legislation, the dominance of status over contract. The GAAR is the apogee of the government’s efforts to ensure the dominance of status over contracts.

Reference Maine HS (1876) Ancient law: its connection with the early history of society, and its relation to modern ideas. John Murray, London

Chapter 2

Overview of Indian Legislation Regarding International Taxation

The Indian law on international taxation is complex even though the foundations of the Indian legislative tax design are actually quite simple. The complexity in Indian tax design, much like the complexity in tax systems across the world, is a response to tax avoidance techniques. The starting point for Indian international tax issues is the same for domestic tax issues: what is charged to tax? Section 5 of the ITA addresses this question sequentially. To begin with, income for tax purposes is defined as total income that arises from any source. Income in the ITA has been envisaged in the widest possible manner, and has in fact not been defined in any restrictive sense at all. The definitional section of the ITA provides an inclusive definition of income and extends the idea of income to, at last count, eighteen categories including perquisites, lottery earnings, crossword puzzle winnings and capital gains.1 However, one of peculiarities of the Indian tax system, which it inherited from the United Kingdom, is its insistence on income emanating from a source. The revenue has to be able to tax income within one of the five heads or schedules of income mentioned in the ITA (commonly referred to as the scheduler system of taxation); otherwise the income is not subject to taxation. This complication is the result of Section 5 of the ITA providing that income for the purposes of taxation is total income. Total income has been defined in Section 2(45) of the ITA as the total amount of income referred to in Section 5, computed in the manner laid down in the ITA. The manner of computation of the income is found in Section 14 of the ITA which provides that all income shall, for the purposes of charge of income tax and computation of total income, be classified under the following heads of income: salaries, income from house property, profits or gains of business or profession, capital gains, and income from other sources. Income from other sources is a residual category that catches income that does not fall within the first four categories.

1 Section

2, ITA.

© The Author(s), under exclusive licence to Springer Nature India Private Limited 2020 N. Nuggehalli, International Taxation, SpringerBriefs in Law, https://doi.org/10.1007/978-81-322-3670-2_2

5

6

2 Overview of Indian Legislation Regarding International Taxation

Thus the revenue has to jump through two hoops; first they have to prove what you have earned is income, and then, they have to prove that your income comes within one of the five heads of income mentioned in the Act. Why are things so complicated? The best answer—though not necessarily the most satisfactory—is to point to the ITA’s British pedigree. The concept of income being divided into various categories or heads is of British vintage and continues to influence the basic structure of Indian tax law. There is no comparative categorization in the United States where if a certain item of profit or earning falls within the definition of income, it is taxable and that is the end of the matter. In the early days of Indian tax law, the second hoop, i.e. pigeonholing income into a “head of income” used to result in some intrepid citizens escaping taxation altogether by arguing that what they received could not be brought within any head of income. But over the years, the revenue has had time to wise up to this argument. The scope of each head of income has been expanded and clarified ad nauseam and additional heads of income (example: capital gain, discussed in a later chapter) have been added, thus diluting the heads of income hurdle to some extent. However the scheduler system continues to play a role in international tax controversies from time to time and will be discussed in more detail later in this book. Section 5 makes taxable income that arises from one of the enumerated sources. It is the requirement that income arise from someplace that brings a territorial nexus element to Indian taxation. India makes an item of income subject to Indian taxation if either it arises with respect to an Indian resident (regardless of where in the world it arises) or if it accrues in India (regardless of the residency of the person with respect to which the income arises). The territorial nexus that India needs in order to assert a jurisdiction to tax income is established in two different ways. India asserts its right to tax income either because a person who is resident in Indian territory has earned income or an item of income has arisen in Indian territory. To use common international tax parlance, India asserts its tax nexus based on both residence and source factors. The first criterion implicates the idea of a resident, and the ITA in Section 6 contains rules for defining these terms. There are rules for defining residency for various kinds of taxpayers i.e. individuals, companies, partnerships etc. The residence rules for companies are most significant, and have been amended recently. Under Section 6(3) of the Act, a company incorporated in India is always considered as resident in India. This is simple enough and follows the trend worldwide. For companies that are not incorporated in India, the rules, were until recently, curiously in favour of treating such companies as non-resident. Accordingly to the rules in force until 2016, a company incorporated outside India was considered resident in India only if the control and management of its affairs was situated wholly in India. The residency rules for companies were fertile ground for tax planning structures by virtue of which companies could claim to be non resident even though their affairs were being managed from India. A successful claim of non residency by a company would entitle that company to shield its non India based income from Indian taxation. From April 1, 2016, the residency rules for companies have been modified by the

2 Overview of Indian Legislation Regarding International Taxation

7

government. A company is considered resident in India if it is incorporated in India or if its place of effective management is in India. The place of effective management refers to the place where the key management and commercial decisions that are necessary for the conduct of business of an entity are, in substance taken.2 The place of effective management test, unlike the incorporation test, introduces some uncertainty in the identification of corporate residency. The test seeks to identify the place where key management decisions are, in substance, taken. The Central Board of Direct Taxes (CBDT) has issued guidelines in an effort to mitigate the uncertainty in this area (CBDT 2017). The Indian tax system aims to tax accretions to wealth that have a source in India. Normally a simple source test would sacrifice certainty at the altar of simplicity because in many situations even the slightest association with India would make an accretion to wealth taxable. For example, a person exporting products from the United Kingdom to India would be taxable because there is a source in India (the paying customer) even if the product was manufactured and financed wholly and exclusively in the United Kingdom. One might say that there is indeed an Indian source here because there is an Indian customer paying for the products being imported. The source debate is not settled by any linguistic usage and therefore could encompass export sales from a person from one country to another country. However until there are further rules to clarify the reach of sourcing rules, there would be continuing uncertainty. One cannot take it for granted that every country would prefer the widest possible reach of the sourcing rules. It might be the case that India will not like to assert its right to tax in the export example given above because exports of its citizens might become subject to like claims from other countries. It is in this context that India, like many other countries, has made additional rules clarifying the ambit and reach of source rules. The Indian tax system does not have a uniform source rule for all types of accretions to wealth. The ITA, presumably in order to redress situations where taxpayers would try to shift income out of India, introduces the legal fiction of income “deemed to arise” in India in Section 9. Like all good legal fictions, this legal fiction has also spawned a variety of interesting legal issues. The ITA creates the legal fiction in a spectacularly sweeping fashion: income is deemed to arise in India if it arises from any business connection in India, or from any property in India, or from any asset or source of income in India or from the transfer of a capital asset situate in India.3 One wonders why some of these cases need to come under a deeming provision. For example, one can say with certainty that income from a sale of a capital asset situated in India (say a house for example) should be considered as “arising” in India.4 However, the open ended wording in Section 9 does suggest that there would be a 2 Section

6(3), Explanation. 9 (1)(i), ITA. 4 Despite India’s assertion of a source based taxation, why should a non-resident be worried about the Indian government going after him for unpaid taxes if his property is situated wholly outside India? In other words, how can India exercise extra-territorial jurisdiction in taxing the income of a non-resident? Needless to say, the Act has a mechanism that solves—at least partially—the 3 Section

8

2 Overview of Indian Legislation Regarding International Taxation

number of instances of income that might have only a tenuous connection with India and yet might be tagged as Indian source income because of the deeming provision in Section 9. It follows from the analysis above that for some kinds of accretions, the ITA requires that the activity that generated the accretion of wealth have a business connection with India, a term which has commonly been taken to require some durable physical presence in India.5 In the main, such accretions of wealth are what are called active business income, i.e. income that one earns when one actually does something involving a durable or substantial physical presence in the host country, such as a case of manufacturing activity. For other kinds of more intangible accretions i.e. accretions that do not require a durable or substantial physical presence on the part of the income earners, the ITA requires that such accretions are a result of money paid out by Indian residents. The two most common examples of the latter are income from royalty payments and income from the provision of technical services.6 Since many countries, of which India is one example, tax on the basis of both source and residence, it is inevitable that the same item of income would in some cases be taxed by two countries. For example, an Indian resident creditor earning interest income from his U.K. debtor will be taxed in India on a residence basis and in the United Kingdom on a source basis. Similarly, a U.K. resident creditor earning interest income from his Indian debtor will be taxed in India on a source basis and in the United Kingdom on a residence basis. The same item of income, i.e. interest income, is being taxed by two countries. How does one solve the problem of double taxation? For example, what does one do with foreign companies who are investing in India, and are therefore likely to be taxed both in India and in their home countries? Should the Indian revenue exempt the income of foreign companies that is earned in India and thus create an exception to the source rule? Similarly, how must the Indian revenue tax Indian companies earning income abroad? If the revenue taxes all of their income, is it not unfair since these companies are also being taxed in the other jurisdictions in which they have decided to invest? If the revenue believes it is unfair for India to impose in full an income tax on these companies, what is the solution? Should India refrain from imposing an income tax on an Indian company’s foreign income or should it relieve the tax burden on the Indian company in some other manner? As one can see, these questions are not resolved by the domestic tax system and need special rules. These special rules are provided under the ITA as well as under double tax avoidance agreements (DTAAs) that India has signed with severel countries. For the past several decades, under the authority of Section 90 of the ITA, India has entered into several double tax avoidance agreements. These double tax avoidance agreements problem of taxing the income of non-residents in India in circumstances where the non-resident and its property may be outside India’s territorial reach. Section 161 of the ITA enables the government to tax persons situated in India as agents of the non-resident. A person is considered as an agent of a non-resident if he has a certain relationship with the non-resident such as having a “business connection” with the non-resident or being employed by him etc. 5 Section 9(1)(i) read with Explanation 1, 2, 2A and 3. 6 Section 9(1)(vi) and Section 9(1)(vii).

2 Overview of Indian Legislation Regarding International Taxation

9

are an interesting breed of international instruments. First, unlike many international instruments, the DTAAs are bilateral, not multilateral. There is some murmur of a multi-lateral treaty but it remains a murmur; it is unlikely that there would be any kind of WTO-like multilateral tax treaty in the near future, although, as will be analysed further in Chap. 7, there is an incipient international multilateralism in tax law (termed the Multilateral Instrument) occurring under the aegis of the OECD.7 Second, unlike most international instruments, it can be argued that the DTAAs in effect replace the domestic tax system applicable to the corporations subject to it. This important legal effect—an alternate rather than an exception to the domestic tax system—can lead to interpretational problems as described below. Section 90 of the ITA is the primary legislative provision setting out the relationship between the ITA and the DTAAs. The section raises several complex issues regarding the relationship between DTAAs and domestic law. Let’s begin with the most straightforward one. Section 90(1)(a) enables the government to relieve double taxation both in circumstances where the same item of income is taxed in two countries as well as when the same item of income is chargeable to tax in two countries.8 This sub section addresses a situation where the same item of income is being subject to tax in two countries. The formula for relief provided under DTAAs entered into by India is to provide a deduction against the Indian tax of the amount of foreign tax paid up to a ceiling. The ceiling is usually the Indian tax paid or payable on the double taxed income. This method of providing relief for double taxation is called the deduction method. An alternative to the deduction method is the exemption method, which is provided for in Section 90(1)(b). Section 90(1)(b) allows for India to enter into DTAAs that will eliminate the possibility of double taxation with respect to certain items of income earned by taxpayers.9 For example, India and the United Kingdom have relieved double taxation in the case of business income by providing under the India U.K. DTAA that a U.K. resident enterprise would be taxable in India only if it has a permanent establishment in India.10 Notice that both the deduction and the exemption method might be applicable to the same kind of income. For instance, it is not the case that the double taxation of all of the business income earned by a foreign enterprise in a state is eliminated. Under the India U.K. DTAA, if UK resident enterprise has a permanent establishment in India, it’s business income will be subject to taxation

7 See

Nuggehalli (2019) on the OECD BEPS projects’ first steps in public international tax law. to Section 90(1)(a)(i), the Central Government may enter into an agreement with the Government of any country outside India for the granting of relief in respect of income on which have been paid both income-tax under this Act and income-tax in that country. Pursuant to Section 90(1)(a)(ii), the Central Government may enter into an agreement with the Government of any country outside India for the granting of relief in respect of income-tax chargeable under this Act and under the corresponding law in force in that country. 9 Section 90(1)(b) provides that the Central Government may enter into an agreement with the Government of any country outside India for the avoidance of double taxation of income under this Act and under the corresponding law in force in that country. 10 Article 5, India U.K. DTAA. Permanent establishments are discussed in Chap. 4. 8 Pursuant

10

2 Overview of Indian Legislation Regarding International Taxation

by India as well as the United Kingdom, in which case Section 90(1)(a) and the consequent DTAA provisions would come into play. Curiously, Section 90 (1)(a) provides for relief with respect to not only income that is taxed in two countries but also income that is chargeable to tax in two countries.11 The case of Wipro Ltd versus DCIT illustrates the force of this sub-section.12 An Indian company earned income from export of software produced in India. In India its income was exempt under Section 10A. The software exported to the United States was taxed in the United States. The issue was whether the US income tax paid on the software exported by Wipro could be allowed as a credit against the Indian tax payable by Wipro on its other activities (activities other than export since the export activity was not subject to Indian tax because of the Section 10A exemption). The revenue argued that the taxpayers could not avail of the double tax credit because there was no Indian tax paid on the software export. The Karnataka High Court decided that pursuant to Section 90(1)(a)(ii) read with Article 25(2)(a), it was sufficient for the double tax credit to be generated if Indian tax was chargeable on the software exports, even though it was exempted under Section 10A.13 In some cases, a certain item of income would be subject to taxation in India and in another country but the relevant DTAA would impose an upper limit on the tax that one country can impose. Usually this upper limit would be imposed on the country in which the income arises. For example, in most of the DTAAs entered into by India, a tax on interest income can be imposed without any limit by the residence state i.e. the state in which the person receiving the interest income resides. Many of the DTAAs also provide that the source state, i.e. usually the state in which the payor of the interests is resident, can also impose an income tax on the interest income but the income tax imposed is subject to an upper limit.14 In this context, the meaning of an income tax can become a subject of controversy. Section 90(1)(a) provides for relief with respect to income taxes paid to or chargeable in two countries on the same income but does not elucidate further the scope of an income tax. The scope of what is meant by an income tax is left to the definitions provided in the respective DTAAs. The case of DIC Asia Pacific Pte Ltd. versus ADIT illustrates the interpretational issues that can arise even with respect to the very idea of what is meant by an income tax.15 The India Singapore DTAA provides that with respect to interest income and royalty income earned by a non resident, the source state, i.e. the state from which the interest and royalties arose, can impose an income tax on the interest income only up to a maximum rate of 11 Pursuant to Section 90(1)(a)(ii), the Central Government may enter into an agreement with the Government of any country outside India for the granting of relief in respect of income-tax chargeable under this Act and under the corresponding law in force in that country. 12 2015 Indlaw KAR 9678. 13 Ibid, para 59. 14 For example, Article 12 of the U.K. India DTAA provides that while the resident state can tax interest income (without any limits), the source state can tax interest income up to a maximum of 15%, subject to some exceptions. 15 [2012] 18 ITR (Trib) 358 (Kolkata).

2 Overview of Indian Legislation Regarding International Taxation

11

15% and 10% respectively.16 An Indian resident had paid out interest to a Singapore resident enterprise. The question that arose in this case was whether the 10% upper limit included the surcharge and the education cess imposed under the Finance Act in India. The India Singapore DTAA defined income for purposes of the Indian tax as including a surcharge but did not mention a cess. The Kolkata Appellate Tribunal decided, based on an analysis of the relevant Finance Act, that a cess was in the nature of a surcharge and therefore the 10% upper limit included the education cess.17 How does the DTAA trump domestic statutory law in tax matters? Section 90(2) states that the ITA will prevail over a DTAA to the extent that the ITA is more beneficial to the taxpayers. This provision suggests that in the normal course, the ITA would be subordinated to the DTAA. Unlike in some countries such as the United Kingdom, the ITA does not explicitly subordinate the domestic statutory tax provisions to the DTAAs. Yet, both the courts and the revenue have steadfastly maintained that the DTAA provisions would prevail over the statutory provisions. One would expect some arcane international law reasons for DTAAs triumphing over statutory provisions. But generally, the courts have tended to use more practical reasons for their decisions in this area. As the Calcutta High Court put it, DTAAs would prevail over the provisions of the Act; “otherwise, there was no point in entering into any agreement for avoidance of double taxation.”18 The usual route the courts use to arrive at the supremacy of the DTAAs over the Act goes like this: The charging provision in the Act—Section 4—is expressly made subject to the provisions of the Act. The same is the case with Section 5 which defines total income of residents and non-residents. The provisions of the Act include Section 90 which is the provision enabling the government to enter into DTAAs. Ergo, the Act is triumphed by DTAAs. If, after the inexorable movement of the preceding logic, you may still have any misgivings, please purge them; the reign of DTAAs is act of faith in Indian tax law today and it delivers the certainty in tax matters for multinationals that is so lacking in other areas of domestic tax law. In the chapters below on permanent establishment and royalties, it will be noticed that DTAAs play the dominant role in setting the standards of taxation and the role of Indian domestic law is reduced to that extent (the opposite is true of the domain of capital gains where the domestic law continues to be dominant). The idea that DTAAs must trump the ITA does not mean that the DTAA can result in an imposition of a tax liability when there can be no tax liability fastened under the ITA. This issue came to the fore in Sumitomo Mitsui Banking versus DDIT.19 The Indian branch of a Japanese bank borrowed money from its head office (based in Japan) and paid interest to its head office in return. The Indian taxability of the interest paid was in question. The Japanese bank argued that it could not be taxed because under the ITA, by virtue of principles set out in the well known Kikabai Premchand case, no one can be taxed

16 India-Singapore

DTAA, Article 11 (interest) and Article 12 (Royalty). para 10. 18 Commissioner of Income-Tax versus Davy Ashmore India Ltd., 190 ITR 626, para 8. 19 2012 Indlaw ITAT 16. 17 Ibid,

12

2 Overview of Indian Legislation Regarding International Taxation

on transactions with themselves.20 The Income Tax Appellate Tribunal (hereinafter ITAT) agreed with the assessee that under domestic tax law, the head office could not be taxed on the interest payments.21 The revenue argued that under Article 7 of the India Japan DTAA, the branch of a foreign corporation is always considered as a separate entity for tax purposes and therefore the taxability of interest payments from the foreign branch must proceed on the separate entity assumption. The ITAT had a preliminary objection to the line of thought: regardless of how the branch is treated under the DTAA, there can be no taxability of the interest payments made by the Indian branch unless the ITA taxes such interest payments. Section 90(2) states that the domestic tax provisions would apply to taxpayers as long as these provisions are more beneficial to the taxpayer. It follows that if the Japanese head office is not subject to tax under the ITA on the interest income received from its branch, the India-Japan DTAA cannot be used to impose a tax on the interest payments.22 In a situation where the Indian branch is receiving an interest payment from the foreign head office, Indian case law supports the imposition of tax liability on such payments.23 The reasoning behind this approach is that the principle that one cannot trade with oneself is a principle that Indian tax law follows only with respect to domestic transactions. In an international context, the principle of self trading does not hold because the permanent establishment is always considered as a separate enterprise in domestic tax law and as per conventional accounting principles. Indian tax law begins by laying down the rule that only the Indian source income of a foreign enterprise would be taxable in India. Section 9 of the ITA has detailed rules on the circumstances in which an income of a foreign enterprise is said to arise in India, and the calculation of Indian source income attributable to a foreign enterprise presupposes the ability to bracket that portion of the foreign enterprise’s income that arises in India. Such a bracketing or allotment is not possible unless the foreign enterprise’s establishment in India is considered (even if notionally or hypothetically) as a separate and independent entity dealing with the foreign enterprise on an arm’s length basis.24 From an Indian tax perspective, there are today two parallel tax systems applicable for accretions to wealth with respect to which the Indian tax authorities assert a territorial nexus to tax. First, there are the double tax avoidance agreements (DTAAs) that India has signed with various countries. The DTAAs would trump any Indian tax provisions to the extent applicable. India has signed DTAAs with most industrialized countries such as the United States, United Kingdom, France, etc. The second legal regime is the Indian Income Tax Act, 1961 that is applicable with respect to the

20 Kikabai

Premchand versus CIT, 1953 Indlaw SC 77. ibid, para 84. 22 Ibid, para 61. 23 Dresdner Bank versus ACIT, (2006) 105 TTJ Mum 149. 24 See paras 22–33 of the Dresdner case (above) for a discussion bearing on the taxability of a branch as a PE. 21 Sumitomo,

2 Overview of Indian Legislation Regarding International Taxation

13

residents of those countries with which India does not have a DTAA. Examples of such countries are Cayman Islands and Jersey. The double tax legal regime is an open legal regime i.e. it brings in or refers to concepts from the legal regimes of the contracting States. For example, the IndiaUK DTAA states that it applies to the residents of India and the United Kingdom but leaves the definition of the term ‘resident’ to each of the contracting States.25 Similarly, Article 14 of the India UK DTAA allows each country to tax capital gains in accordance with the respective provisions of its domestic law. Such an open legal regime can lead to interpretation problems. First, a contracting state can amend its tax law with respect to the concepts referred to in the tax treaty in such a way that the benefits under the tax treaty are adversely affected. The taxpayers have argued that the rights under the treaty must remain as they were on the date the treaty came into force and cannot be affected by subsequent amendments to domestic tax law. This issue has not yet been resolved conclusively in India; the Indian Supreme Court is likely to rule on his issue at some point of time. Another related complication, which also remains unresolved, relates to the taxpayer rights under tax treaties that are impacted by retrospective amendments to domestic tax law. The Indian domestic tax regime had seven important features that are replicated, allowed or assumed in its tax treaty regime. The ITA makes a distinction between income of a capital nature and that of a revenue nature,26 the taxation of income is schedular in nature,27 the tax system makes a distinction between residents and non-residents,28 the tax system divides accretion to wealth in terms of domestic and foreign source,29 the tax is paid by different tax units with the main distinction between an individual and the corporation,30 wealth accretion is generally taxed on

25 Article 4, Agreement for avoidance of double taxation and prevention of fiscal evasion between India and the United Kingdom of Great Britain and Northern Ireland (Hereafter the India-UK DTAA). 26 Under ITA Section 4, the term “income” has been defined in the Act in Section 2(24). Income has been defined as including a number of enumerated items such as profits and gains, dividend, the value of perquisites, capital gains, winnings from lotteries, and sums received under insurance policies. However, a receipt of a capital nature is not taxable. For example, a sum of money received for transferring carbon credits was considered as a capital receipt and therefore outside the purview of the ITA ( Subhash Kabini Power versus CIT, 2015 (1) TMI 646, overruled by the Finance Act, 2017 through the introduction of Section 115 BBG. 27 The schedular system divides income according to the source of income. Therefore if an income is from a certain source, say the letting of property, it is taxable under the schedule most directly related to that source (income from house property) even if there are other schedules under which this income can potentially be classified (Raheja IT Park vs. CIT, 2014 11 TMI 349). 28 As provided in Section 5 of the Income Tax Act, 1961, residents are taxed on their income without any regard to where it is earned, i.e. without any regard to whether it is earned in India or elsewhere. On the other hand, non-residents are taxed only on income earned in India. 29 Section 9, ITA. 30 The ITA identifies different kinds of taxpayers. Section 5 imposes an income tax on a ‘person’. Section 2(31) defines a person to include an individual, a Hindu Undivided Family, a company,

14

2 Overview of Indian Legislation Regarding International Taxation

a net rather than a gross basis,31 the calculation of the accretion to wealth for tax purposes is dependent on underlying accounting principles,32 and the ITA has special provisions to deal with abuses of the tax system.33 In 2001, with effect from 2002, India introduced transfer pricing regulation in order to regulate the value of transactions between entities in a multinational group. Section 92 of the ITA states that ‘any income arising from an international transaction shall be computed having regard to the arm’s length price’. Transfer pricing is the most contentious area in international tax law and is discussed at length in Chap. 8. Moving beyond the issue of taxable nexus with regard to foreign persons in India, Indian tax law recognises the problem of collecting taxes from foreign persons, even if taxable nexus is established. Towards this objective, the ITA has an elaborate set of mechanisms to impose collection responsibilities on Indian residents that are in a commercial relationship with foreign taxable persons. The typical mechanism, followed by administrations worldwide, is to impose a withholding requirement on certain payments to foreign persons. Following this trend, the ITA has imposed a withholding requirement on Indian residents when they pay interest, royalties, capital gains or dividends to foreign persons.34 Breach of withholding obligations has serious consequences-the Indian resident would be liable for the tax that not been withheld and would also be subject to penalties and interest.35 Non-withholding on payments to foreign persons will also lead to such payments being made ineligible for deductions under the ITA.36 It should be noted here, although it is a fairly obvious point, that the Supreme Court had made it clear that if a foreign person does not owe any income on a payment received by it, there would be no withholding obligation on the Indian resident making the payment to the foreign person.37 With respect to business income, an Indian resident in a business connection with a foreign person is deemed to be a representative assessee of the foreign person and will be liable to be taxed to the same extent and in the same manner as the foreign person.38 With respect to capital gains, the ITA imposes a withholding obligation on the purchase of property; if there is no withholding, the purchaser would be liable for the tax on the capital gains in addition to interest and penalties.39 a firm, an association of persons or a body of individuals, whether incorporated or not, a local authority, and every artificial juridical person, not falling within any of the preceding categories. 31 This is particularly important for business income. Section 37(1) of the ITA states that any expenditure (not being expenditure of the nature described in Sections. 30–36 and not being in the nature of capital expenditure or personal expenses of the assessee), laid out or expended wholly and exclusively for the purposes of the business or profession shall be allowed in computing the income chargeable under the head ‘Profits and gains of business or profession’. 32 Section 145, Income Tax Act, 1961. 33 For example, Chapter X and Chapter X-A, Income Tax Act, 1961. 34 Section 195, ITA. 35 See Sections 201, 220 and 221 of the ITA. 36 Section 40 of the ITA. 37 GE India Technology Centre (P) Ltd. versus CIT [2010] 327 ITR 456 (SC). 38 Section 160 read with Section 163 of the ITA. 39 See Sections 195, 201, 220 and 221 of the ITA.

2 Overview of Indian Legislation Regarding International Taxation

15

The survey of international tax principles in this chapter makes it clear that in reality the phrase international tax is somewhat of a misnomer if it tends to conjure in our minds some kind of a supra-national law that is applied globally; international tax usually refers to the application of a country’s tax principles (rather than global tax principles) to transactions that have an international dimension, usually due to that country’s companies investing overseas or foreign companies investing in that country. In the past fifty years, more and more countries have entered into bilateral “double tax avoidance” agreements (DTAAs) with each other. Since these agreements have generally standard provisions, decisions by domestic tribunals across the world can perhaps be looked at as an “international law” of tax. Courts in various countries frequently refer to each other’s decisions in interpreting double tax agreements and so, in a manner of speaking, a loose melange of common tax law principles is beginning to take shape. The advent of the OECD BEPS process and the emergence of the Multilateral Tax Instrument (discussed in Chapt. 7) will accelerate the emergence of a common set of tax principles worldwide.

References CBDT (2017) Circular No. 06 of 2017, Guiding Principles for determination of Place of Effective Management (POEM) of a Company, 24th January, 2017, available at https://www. incometaxindia.gov.in/news/circular06_2017.pdf Nuggehalli N (2019, January 21) BEPS and the dawn of public international tax law [Blog post]. Retrieved from https://ijiel.in/blog/f/beps-and-the-dawn-of-public-international-tax-law

Chapter 3

Contracts, Status and the Judiciary

The English Judiciary and Tax Planning They say an Englishman’s home is his castle. The moat that secures his castle against indiscriminate state appropriation is a constitutional principle of impressive vintage: No tax shall be imposed without the authority of a statute. This principle has lead to what some might consider an unhealthy obsession with statutory language. The idea that a statute’s language has to account for any tax imposed was taken by judges initially to mean only taxes that come within the literal language of the statute would be allowed. As one would have expected, some taxpayers took their cues from this and concocted schemes that stretched the statutory language and eventually the credulity of judges. The judicial mood soured. Some judges were scathing in their assessment of taxpayers who enacted tax saving schemes that gamed the tax system. Consider this classic piece of judicial invective in Lord Templeman’s words: The facts set out in the case stated by the special commissioners demonstrate yet another circular game in which the taxpayer and a few hired performers act out a play; nothing happens save that the Houdini taxpayer appears to escape from the manacles of tax. The game is recognisable by four rules. First, the play is devised and scripted prior to performance. Secondly, real money and real documents are circulated and exchanged. Thirdly, the money is returned by the end of the performance. Fourthly, the financial position of the actors is the same at the end as it was in the beginning, save that the taxpayer in the course of the performance pays the hired actors for their services.1

1 [1979]

1 WLR 974, 97. This passage was quoted by Lord Walker who described this passage as ‘vintage prose’. See Lord (2004, p. 413). © The Author(s), under exclusive licence to Springer Nature India Private Limited 2020 N. Nuggehalli, International Taxation, SpringerBriefs in Law, https://doi.org/10.1007/978-81-322-3670-2_3

17

18

3 Contracts, Status and the Judiciary

Despite Lord Templeman’s lamentations, the judiciary in England tends by and large to acknowledge that the tax status of taxpayers is impacted greatly by the contractual relationships entered into by taxpayers. In fact, the first celebrated case on tax avoidance is also the first case to bring out the special connection between contracts and taxation. In IRC versus Duke of Westminster,2 the Duke of Westminster paid wages to a retinue of staff including his gardener Frank Allman. The tax law as that time in the UK did not allow the Duke to deduct the wages paid to his staff from his taxable income. On the other hand, the law allowed a deduction of annuities paid to other people.3 The Duke entered into a simple tax avoidance scheme. He bestowed annuities on his employees in the form of ‘executed deeds of covenants.’4 The covenant executed in favour of his gardener Frank Allman was a typical example. The covenants promised to pay the gardener a specified weekly sum for a period of seven years in consideration of past services. A letter of explanation was sent to the gardener informing him that there was nothing to prevent the gardener from claiming full remuneration for his work but it would be expected “in practice” that he would be content with the annuities being paid to him.5 The annuities being paid to the gardener were in fact the equivalent to the wages he was receiving earlier (the letter to the gardener indicated that even if the annuities fell short of the gardener’s wages, the Duke would consider making up the difference).6 When the Duke went ahead and deducted the annuities from his taxable income, the UK revenue challenged the deduction in the courts. The revenue claimed that in substance the Duke was attempting to deduct the otherwise non-deductible wages paid to his employees. The revenue did not claim that the deed of covenant issued in favour of the gardener was a sham. Indeed, they could not make such a claim. There is a high threshold of proof for sham cases in the common law. The revenue would have had to prove that the drafters of the documents never intended to abide by the terms of the document and that the legal rights and duties created by the document were illusionary and merely a cover for some other reality. The revenue knew they would not prove the Dukes’ covenants to be a sham. The Duke was perfectly willing to recognize and give effect to the legal rights of the gardener to his annuity. Therefore, the revenue went on another path: the notion of substance versus form. In form, the revenue admitted that the deeds of covenants were annuities. But the substance of the transaction, the revenue contended, was to provide wages to the gardener. The revenue’s position was understandable. The annuities were equivalent in amount to the wages (with the Duke expected to make up any difference). True, according to the letter, the gardener could have also asked for wages, but considering his position and his relationship with the Duke, the gardener’s so called right to wages

2 [1936]

AC 1 (HL).

3 Annuities and wages are both periodic sums of money but while wages are paid on a current basis,

annuities are paid for past services rendered. AC 1 (HL) 2. 5 [1936] AC 1 (HL) 12. 6 See Footnote 5. 4 [1936]

The English Judiciary and Tax Planning

19

was so unlikely that the very notion of him standing up to the Duke and demanding his wages in addition to the annuities was laughable. Yet, the House of Lords was extremely reluctant to ignore the legal rights and duties created by the contract in this case. The Duke’s covenants created a legal reality that could not be ignored by the courts. This ability of a citizen to change his tax liability by creating different kinds of contractual rights and duties is a facet of the modern state that has always been acknowledged and the House of Lords was reluctant to disturb the status quo. Lord Tomlin made a remark that was not only to be quoted far into the future by tax lawyers and judges alike but also prove to be a panacea for the tax avoidance industry: Every man is entitled, if he can, to order his affairs so that the tax attaching under the appropriate Acts is less than it otherwise would be. If he succeeds in ordering them so as to secure this result, then however unappreciative the Commissioners of Inland Revenue or his fellow taxpayers may be of his ingenuity, he cannot be compelled to pay an increased tax. This so-called doctrine of “the substance” seems to be nothing more than an attempt to make a man pay notwithstanding that he has so ordered his affairs that the amount of tax sought from him is not legally claimable.7

The House of Lords came to the conclusion that the legal rights and duties created by the annuities contracts had to be respected, thus allowing the Duke to deduct the annuities from his income. The gardener, you should note, was indifferent to the difference between annuities and wages as under both the contractual regimes he was getting his due. But the Duke, by using different legal documents to channel the gardener’s dues, improved his tax circumstances. The Westminster case came to be widely regarded as a triumph of contract over status, and matters have remained much the same in the area of tax avoidance despite the supposedly more aggressive approach of the House of Lords in Ramsay versus IRC.8 In Ramsay, the taxpayer had a capital gain with regard to which the taxpayer entered into a tax avoidance scheme. Each of the transactions in the scheme was a genuine legal transaction, i.e. there was genuine legal documentation for each of the transactions. But the transactions essentially cancelled themselves out while also providing a useful loss to the taxpayer. The revenue could not argue that the tax avoidance scheme was a sham; it most certainly was not. However, they argued that the transaction producing loss and the transaction producing gain should be analyzed together and not separately. Treated separately, the taxpayer was right in claiming a loss; treated together the taxpayer had nothing to gain since there would be no transaction to look at since the two transactions cancelled each other out. The House of Lords chose to approach the scheme in the following fashion: Given that the document or transaction is genuine, the court cannot go behind it to some supposed substance. This is the well-known principle of IRC v Duke of Westminster. This is a cardinal principle but it must not be overstated or over-extended. While obliging the court to accept documents or transactions found to be genuine as such, it does not compel the court to look at a document or a transaction in blinkers, isolated from any context to which 7 [1936] 8 [1982]

AC 1, (HL) 19–20 (Lord Tomlin). AC 300 (HL).

20

3 Contracts, Status and the Judiciary it properly belongs. If it can be seen that a document or a transaction was intended to have effect as part of a nexus or series of transactions, or as an ingredient of a wider transaction intended as a whole, there is nothing in the document to prevent it being so regarded; to do so is not to prefer form to substance, or substance to form. It is the task of the court to ascertain the legal nature of any transactions to which it is sought to attach a tax or a tax consequence and if that emerges from a series or a combination of transactions, intended to operate as such, it is that series or combination which may be so regarded.9

The Lords in Ramsay were careful not to be seen to be overruling the Duke of Westminster Principle, and indeed, the Ramsay principle when applied to the Westminster case would not have changed the decision in that case presumably because there was only one transaction instead of a series of transactions. As can be seen from the Ramsay quote above, the court was also careful not to lay down some version of a substance over form principle, a principle so beloved of American revenue authorities. Perhaps the Lords were perhaps reluctant to espouse such a principle for if left unchecked, this principle would be unjust to the taxpayers because it makes it impossible for the taxpayers to contractually structure their transactions in such a way that it has the same or nearly the same economic effect as a taxable transaction while falling outside the terms of the taxing statute. In other words, liberal use of the “substance over form” doctrine could take away the ability of taxpayer to minimize its taxes through strategic contracts. However, there was a judicial tendency post Ramsay to consider Ramsay as some kind of a judge made rule, and the elaborate manner in which the Ramsay rule was discussed gave rise to the suspicion that this rule existed apart from and in addition to any canon of statutory interpretation. For example, consider Furniss versus Dawson.10 In Furniss, the taxpayer was interested in selling the shares of his company. In order to avoid capital gains tax, the taxpayer first transferred the shares of his company to an Isle of Man company (in return for all the shares in the Isle of Man company) and then made the Isle of Man company sell his company’s shares to the ultimate buyer. The court, using Ramsay principles, disregarded the transfer of shares to the Isle of Man company and looked at the transaction as a straight sale of shares from the taxpayer to the ultimate buyer. In disregarding the role of the Isle of Man company, the Ramsay doctrine was applied in the following manner by Lord Brighton. Notice the precision of this rule, an aspect that was picked on by academic commentators later. First, there must be a pre- ordained series of transactions; or, if one likes, one single composite transaction. This composite transaction may or may not include the achievement of a legitimate commercial (i.e. business) end…Secondly, there must be steps inserted which have no commercial (business) purpose apart from the avoidance of a liability to tax - not “no business effect”. If those two ingredients exist, the inserted steps are to be disregarded for fiscal purposes. The court must then look at the end result. Precisely how the end result will be taxed will depend on the terms of the taxing statute sought to be applied.11

9 [1982]

AC 300 (HL) 323–324. A.C. 474. 11 Ibid, 527. 10 [1984]

The English Judiciary and Tax Planning

21

The Lords’ balancing act in Ramsay was a commendable feat of judicial dexterity except for one problem: the artificiality of tax legislation and tax treaties. This was a particularly acute problem when it came to taking advantage of special tax breaks or concessions instituted by statutes or double tax treaties. Taxpayers were able to convince the courts in cases arising after Ramsay that the series of transactions undertaken by them had in fact exactly the legal effect envisaged by the relevant tax legislation or tax treaty thus entitling them to tax breaks even if the economics of the transaction were not what the tax legislation was aimed at. This problem was at its most acute in Barclays Mercantile Business Finance Limited versus Mawson.12 The UK revenue alleged in Barclays that the taxpayer took advantage of tax legislation providing liberal depreciation deductions to businessmen willing to invest in capital equipment. The taxpayer, which had a banking and equipment leasing business, bought capital equipment at market value from a seller to whom it leased the same equipment at a rate slightly less than the (discounted) market value of the capital equipment. The difference was the fee the seller received for entering into the transaction. The taxpayer began using depreciation deductions arising by virtue of the taxpayer’s ownership of the capital equipment. The seller put the cash it received from the taxpayer back into the taxpayer’s banking business as a “security” for the rent payable on the lease. According to the UK revenue, when the dust cleared, the taxpayer was not out of cash since it had received the bulk of the money back in the form of the security deposit, but it was now entitled to liberal depreciation deductions, and the seller had received a small fee for its part in the transaction. The House of Lords in Barclays gave a decision in favour of the taxpayer in this case. The Lords were satisfied that the statute’s purpose had been served as long as there had been a capital investment (in this case, the lessor buying the property and then leasing it). The House of Lords stated: ‘if the lessee chooses to make arrangements, even as a preordained part of the transaction for the sale and lease back, which result in the bulk of the purchase price being irrevocably committed to paying the rent, that is no concern of the lessor’.13 The Barclays decision can be seen as an endorsement of the proposition that taxpayers can, except in certain limited circumstances, create their tax status through their contracts. The Barclays decision also raises an interesting question of statutory interpretation. The House of Lords in Barclays opined that the Ramsay case ‘did not introduce a new doctrine operating within the special field of revenue statutes’. On the contrary, as Lord Steyn observed in McGuckian, [1997] 1 WLR 991, 999 it rescued tax law from being “some island of literal interpretation” and brought it within generally applicable principles.14 Prof. Judith Freedman has challenged this interpretation of the Ramsay principle. According to Freedman, Ramsay creates two levels of judicial decision-making: one at the level of deciding the nature of transaction to tax, and another, at the 12 [2004]

UKHL 51, [2005] 1 AC 684. UKHL 51, [2005] 1 AC 684 [42]. 14 Ibid [33]. 13 [2004]

22

3 Contracts, Status and the Judiciary

level of ascertaining the intent of the Parliament in enacting the tax legislation. Ramsay, therefore, asks courts to arrive at a cumulative decision that depends on “this combination of statutory construction and the special analysis of the composite transaction” (Freedman 2007, p. 57). Recalling Lord Brightman’s statement of the Ramsay principle in Furniss, Freedman believes that the “precision with which” the Ramsay rule “is spelt out … does suggest the existence of some kind of judicial rule going beyond statutory construction” (Freedman 2007, p. 60).15 On the other hand, consider Lord Hoffmann’s approach to statutory interpretation. According to Lord Hoffmann, statutory interpretation, properly understood, refers to a purposive construction of statutes, in order to give effect to the intention of the parliament (Hoffmann 2005). Lord Hoffmann refers to the famous US tax case of Helvering versus Gregory16 where the US Court of Appeals purposively interpreted the phrase ‘plan of reorganisation’ in the tax legislation to deny a taxpayer a tax benefit because the taxpayer’s plan of reorganisation was not in pursuant of a purpose endorsed by the legislature. According to Lord Hoffmann, Helvering is the appropriate manner in which to interpret a statute. He argues that purposive interpretation of statutes is actually hampered by legislation that provides for detailed rules for transactions because in such cases judges are forced to give effect to the rules rather than purposively interpret the legislation. Lord Hoffmann’s observation relates to how recent statutes, particularly tax statutes, introduce detailed rules for tax treatment of commercial transactions and therefore present a problem for judicial interpretation of tax statutes. Detailed tax legislation can prevent a judge from adopting decisions that would address economic arguments; the judge is restricted to a limited set of choices due to the detail of the legislative directive. Much depends on the level of detail in the legislation. If the legislation provides a tax deduction for a 10-feet long corrugated steel pipe, then the judge has limited options while implementing the legislative goal through his decision in a particular situation where a taxpayer pays for a 10-feet long corrugated steel pipe. The above analysis of how legislative detail can affect judicial choice is captured in Lord Hoffmann’s argument: Parliament may not be content to describe the economic event which should attract tax because it does not trust the courts to understand such a concept and apply it in a practical way. Instead, it enacts a mass of detailed rules which it is hoped will tie up the taxpayer in a net from which he cannot escape. But sometimes there are holes in the net and the courts find that they cannot plug them by appealing to the economic event which, at a higher level of generality, it appears that Parliament wished to tax. It is one thing to give the statute a purposive construction. It is another to rectify the terms of highly prescriptive legislation 15 Freedman writes about the judicial decision-making in Barclays in the following terms: ‘Had they

believed that the circularity of the scheme was relevant, they would have analysed the transaction as a composite whole. The reason they did not do so related to the nature of the scheme, not the statute. This goes further than pure statutory construction, even of a purposive nature, since the outcome of the application of the statutory words depends upon this special style of transaction analysis and not just a reading of the wording in the statute.’ (Freedman 2007, p. 66). 16 69 F 2nd 809 (1934).

The English Judiciary and Tax Planning

23

in order to include provisions which might have been included but are not actually there. (Hoffmann 2005, p. 205)

While Lord Hoffmann (one of the judges in the Barclays decision) believes that statutory interpretation is nothing but purposive interpretation, Freedman believes that in the guise of statutory interpretation, tax law has moved towards the application of a judge made rule that is not part of but in addition to statutory interpretation. I will present a theory of statutory interpretation in Chap. 9 that will make a case for a wider perspective on interpreting statutes that includes the kind of judicial approaches highlighted by Freedman. Although Barclays can be seen as a victory for tax planning, there are some limits (although difficult to pin down precisely) to the judicial toleration of the exploitation of tax reliefs given for capital expenditure. A case in point is HMRC versus Tower M Cashback.17 The tax avoidance in this case was so blatant that even the popular press took note of the facts and the decision. Here’s how the Guardian reported the decision: The scheme involved a software company called MCashback. As part of the structures, investors put up 25% of the cost of buying software from the company, the other 75% coming from a loan indirectly provided by MCashback itself. The investors then tried to claim a special software-related tax relief for the full 100% of the cash paid for the software, enabling them to get £40 off their income tax bills for every £25 they put up. The Supreme Court ruled that they could only claim the tax relief for the cash that they invested. (Hawkes 2011)

Notice that once again, the issue was, like Barclays, whether the taxpayer had really incurred expenditure by paying or investing in a certain property (software). One way to distinguish this case from Barclays is to notice how in Barclays there was indeed in real terms an expenditure at market rates by Barclays on the pipeline. In Tower M Cashback, the expenditure by the investor on the software was hugely inflated and was well in excess of the market value of the software.

The Indian Judiciary and Tax Planning The Indian case law on tax planning really kicks off with a case that is concerned with sales tax and excise duties rather than income tax provisions. In Mc Dowell & Company Ltd. versus The Commercial Tax Officer,18 the Supreme Court was faced with an ingenious attempt to avoid the payment of a sales tax. The sales tax legislation in the state of Andhra Pradesh imposed a tax on the turnover of companies. The taxpayer was the manufacturer of liquor. Normally the excise tax paid by the taxpayer on the manufacture of the liquor would be part of the price paid by the purchasers and therefore would be part of the taxpayer’s turnover. In order to reduce 17 [2011] 18 [1985]

UKSC 19. 3 SCR 791.

24

3 Contracts, Status and the Judiciary

its turnover for sales tax purposes, the taxpayers asked the buyers to pay the excise duty directly to the government. The taxpayer then claimed that the purchase price minus the excise duty was the turnover for sales tax purposes. The revenue challenged the taxpayer’s attempt to reduce its sales tax burden in this manner. When this issue first came before the Supreme Court, the court actually agreed with the taxpayer and stated that the excise duty paid directly by the purchasers did not go into the common till of the sellers and as a consequence did not become part of their circulating capital (and by extension, their turnover).19 As a result, the taxpayer succeeded in reducing their sales tax burden to the extent of the excise duty paid by them. Immediately after the favourable decision, the government amended the excise rules to make it clear that it was the manufacturer that had to pay the excise duty before liquor was released to the purchasers. Nevertheless, the assessee continued its practice of asking the buyers to pay the excise duty on the liquor directly and argued that the sales tax ought to be paid on the purchase price exclusive of the excise duty.20 The revenue challenged the assessees position and once again the same issue reached the Supreme Court. This time, the Supreme Court refused to give their blessings to this attempt to avoid tax. However, beyond its general denouncement of the evils of tax avoidance, which would be discussed below, the reasons for the Supreme Court’s decision were related to the amended excise rules. Since the amended excise rules imposed the duty to pay the excise tax on the manufacturer, the buyers making the excise payment instead would be considered as doing so for the seller’s benefit. Hence, the consideration for the liquor would include, in addition to the price of the liquor mentioned in the bill, the monetary benefit conferred by the buyer on the seller by paying the excise duty on behalf of the seller. The Supreme Court further made some comments on the judicial approach to tax avoidance, which, at least for a period of time, caused much trepidation for taxpayers and their advisors: In our view the proper way to construe a taxing statute, while considering a device to avoid tax, is not to ask whether the provisions should be construed literally, or liberally, nor whether the transaction is not unreal and not prohibited by the statute, but whether the transaction is a device to avoid tax, and whether the transaction is such that the judicial process may accord its approval of it.21

This statement appears to take the position that there was some discretion in the hands of judge to consider whether a tax avoidance measure was permissible even if 19 Ibid,

811. is unclear from the judgment whether there was an express contract under which the buyers agreed to pay the excise duty on the liquor. The Supreme Court proceeded on the basis that there was an agreement or an arrangement between the seller and the buyer regarding the payment of the excise duty. For example, on p. 817, the court stated ‘We would like to add, that the position is not different, when under a prior agreement, the legal liability of the manufacturer-dealer for payment of excise duty is satisfied by the purchaser by direct payment to the excise authorities or to the state exchequer. 21 Ibid, 809. 20 It

The Indian Judiciary and Tax Planning

25

it was based on a reasonable interpretation of the relevant tax legislation. Although these comments were expressed in a separate opinion of one of the judges (Justice Chinnappa Reddy), the other judges made it known that they approved of the separate opinion.22 When the Supreme Court stated that the key question in tax avoidance cases is whether the court must approve of the tax avoidance, it gave rise to a reasonable apprehension in the minds of taxpayers that the legal rights and duties flowing out of their contractual relationships might well be ignored by the revenue (with the support of the courts) if the courts believed the contracts did not deserve judicial benediction. Quite in what circumstances such a judicial benediction would be forthcoming was not spelt out in McDowell. What was more worrying was that Justice Chinnappa Reddy, in describing the judicial attitude towards tax avoidance, made it clear that the Westminster doctrine is a dead letter law as far as Indian jurisprudence was concerned. He stated that ‘in the very country of its birth, the principle of Westminster has been given a decent burial’ and that ‘no man now can get away with a tax avoidance project with the mere statement that there is nothing illegal about it’.23 The Westminster doctrine was widely seen in the tax advice industry as blessing legitimate attempts to reduce one’s tax liability within the four corners of the law. Therefore a judicial pronouncement sounding the death knell of the Westminster doctrine, combined with the wide amplitude given in Justice Chinnappa Reddy’s judgment to attack tax avoidance measures, led to much anxiety among tax advisers, and raised fears of the judicially determined status of commercial transactions completely overshadowing the contractual intentions of the parties to the transactions. This was, needless to say, a very unsatisfactory state of affairs, which was fortunately remedied to some extent in Union of India versus Azadi Bachao Andolan.24 In Azadi Bachao Andolan (ABA), the petitioners were a civil society group that petitioned the Supreme Court to prevent what they felt was a national outrage: the India-Mauritius DTAA was being used to reduce taxes owed legitimately to the Indian government. The immediate provocation for the petition was a circular issued by the revenue that clarified that in order to obtain any treaty benefits under the IndiaMauritius DTAA, a residency certification from the relevant Mauritius authority was sufficient. This circular raised the hackles of the petitioners because obtaining a residency certificate without also having a substantial physical presence in Mauritius is possible. The petitioners claimed that the circular was going to be used as a tax avoidance mechanism to obtain the benefits of the India-Mauritius DTAA without having any real presence in Mauritius. In an argument directly relying on the Supreme Court’s comments in McDowell, the petitioners claimed that ‘any tax planning which is intended to and results in avoidance of tax must be struck down by the Court.’25 This argument, if accepted 22 Ibid,

824. 807. 24 2003 Indlaw SC 823. 25 Ibid [139]. 23 Ibid,

26

3 Contracts, Status and the Judiciary

by the court, would have sounded the death knell for tax planning and confirmed the absolute and permanent suzerainty of status over contract. Fortunately for investors and their financial advisers, the Supreme Court refused to extend McDowell in the manner asked for by the petitioners. The Supreme Court stated that it would not endorse the ‘extreme view’ of Justice Chinnappa Reddy, which was to cast doubt on any transaction intended to avoid tax.26 The Supreme Court denied that the Westminster doctrine was not valid anymore, as Justice Reddy had asserted. The ABA judgement went through the corpus of English judgements post Westminster and concluded that the Westminster principle continued to hold ground in English law.27 The ABA judgment concluded that tax planning which is otherwise valid in law cannot be treated as ‘non est’ because the scheme adopted by the taxpayer had some underlying tax avoidance motive.28 Apart from granting benediction to tax planning (not tax abuse), the ABA judgment is also noteworthy for its approach to the interpretation of double tax treaties. The Supreme Court made a sharp distinction between domestic legislation and tax treaties.29 It characterised tax treaties as products of commercial negotiations and diplomacy between nations, intended for a number of reasons that go beyond an allocation of tax revenue between the treaty partners.30 Developing countries frequently make use of tax treaties that allow treaty shopping in order to attract much needed capital and technology. In this respect, the role of tax treaties is similar to the role of tax holidays or other tax incentives in attraction foreign investment.31 Given this characterisation of the nature of tax treaties, the Supreme Court was less inclined to treat the idea of post box companies (as one example of treaty shopping) as alarming. It pointed out that Mauritius has been considered as a conduit for foreign investment in India and in fact the Indian government would not have been unaware of the tax benefits that accrued to the residents of Mauritius because of the IndiaMauritius DTAA.32 India intended to and did indeed derive significant economic benefits from the treaty and therefore an interpretation of the treaty provisions must keep this larger commercial picture in mind. To sum, the Supreme Court made it clear that given the commercial context of the tax treaty, it would not interpret the provisions to require the putative residents to provide anything beyond what the plain meaning of the treaty provisions required in order to prove their residential status. ABA established firmly the primacy of contracts in international taxation. Taxpayers could not get away with sham transactions but barring such instances, taxpayers could fashion their transactions any which way they wanted in order to reduce their 26 Ibid

[144]. [143]–[163]. 28 Ibid [164]. 29 Ibid [133]. 30 Ibid [134]–[135]. 31 Ibid [136]. 32 Ibid [138] where the Supreme Court stated ‘Despite the sound and fury of the respondents over the so called ‘abuse’ of ‘treaty shopping’, perhaps, it may have been intended at the time when Indo-Mauritius DTAC was entered into. 27 Ibid

The Indian Judiciary and Tax Planning

27

tax liabilities. In the international tax arena, ABA was not an outlier. Multinational corporations have long taken advantage of the primary of contracts to reduce their international tax liability. As discussed in Chapter six (capital gains), the Supreme Court in the Vodafone case affirmed the ABA position.

References Freedman J (2007) Interpreting tax statutes: tax avoidance and the intention of parliament. 123 LQR 53 Hawkes A (2011, May 11). Pop stars among investors hit by taxman’s victory. Retrieved from https:// www.theguardian.com/business/2011/may/11/revenue-wins-tax-avoidance-scheme-ruling Hoffmann L (2005) ‘Tax Avoidance’. BTR 197 Lord W (2004) Ramsay 25 years on: some reflections on tax avoidance. 120 LQR 412

Chapter 4

Permanent Establishment

The concept of permanent establishment is related to the idea of a state establishing territorial nexus over a non-resident’s business income. The ITA, in section 14, has created a separate category of income called “Income from profits and gains of business or profession”. Section 28 explains the reach of this category of income. Section 28 does not define business income, but section 2(13) of the Act has defined a business as including any trade, commerce, manufacture or any adventure or concern in the nature of trade, commerce or manufacture. This is a fairly comprehensive definition, covering every possible facet of doing work for profit. However there must be a systematic activity with a profit motive for there to be a business.1 1 In

Saroj Kumar v. I.T. Commissioner, AIR 1959 SC 1252, the Supreme Court commented on the meaning of “adventure in the nature of trade”. The taxpayer earned income as the shareholder and director of several limited liability concerns, and also as a partner in an engineering firm. The taxpayer bought and sold a plot of land, and the question was whether the sale transaction resulted in “business income”. In other words, whether the sale transaction has such characteristics as to be an adventure in the nature of trade? The revenue argued that the area of land in question—three quarters of an acre in the suburbs of Calcutta—was large enough to indicate that the assessee would not have intended to buy it for his own use and occupied (and therefore must have intended to resell it). Also the taxpayer had taken a loan to finance his purchase of the land and the revenue argued that he could not possibly paid the outstanding balance on the purchase money (thus indicating that the taxpayer always wanted to sell the plot). The Supreme Court, at para 10 of its judgement, found for the taxpayer. The Court stated that the site was not so large so as to lead to the inference that it could not have been meant to be used by him in the way of his own business or for his own residence. The taxpayer was carrying on an engineering concern, and according to the court, he may have intended, as the taxpayer himself argued, to put up a small workshop on a portion of the land to be acquired, and to build a residential house on the other portion. Another fact was pertinent: at the time he bought the land, the taxpayer’s business was doing well. The court indicated that therefore the taxpayer might have been able to raise an amount of money sufficient to construct a residential house or a workshop on the land. The court also found it pertinent (para 7) that the taxpayer was not in the business of buying and selling land. After considering all these circumstances, the court decided that the sale of the plot was not in pursuance of a venture in the nature of a trade or business. The decision turns on facts; the court looked into the economic circumstances of the taxpayer, the nature of his work and his profession. The guiding principles in this area—mentioned above—only

© The Author(s), under exclusive licence to Springer Nature India Private Limited 2020 N. Nuggehalli, International Taxation, SpringerBriefs in Law, https://doi.org/10.1007/978-81-322-3670-2_4

29

30

4 Permanent Establishment

Business Connection The concept of ‘business connection’ is key to India’s assertion of territorial nexus over a non-resident’s business income. In order to tax the business income of a nonresident, section 9 requires that the business income must arise out of a ‘business connection’ with India. Of all the deemed income categories discussed above, the business connection category has proved the most recalcitrant to sensible interpretation, and has been analysed many times by the Indian judiciary. The term “business connection” is so wide that it had to inevitably undergo judicial refinement. Otherwise, the term, like the accordion, can be stretched to include any and all kinds of commercial dealings that non-residents [i.e. persons including companies that are not resident in India] can have with Indian residents. Say Mr. Jones in American decides to sell some bars of soap to Mr. Patel in Gujarat. Both prior to this and subsequent to this, Jones does not sell any product to an Indian resident. Does this single isolated transaction lead to a business connection in India? Jones and Patel have a business relationship but is this relationship the kind that the Act is seeking to tax? The most detailed explanation of this concept can be found in the Supreme Court case of Commissioner of Income Tax, Punjab v. R. D. Aggarwal & Co.2 Two nonresident companies had hired an Indian representative to procure sales contracts in India. Although the representatives were marketing the products of the non-resident companies, the contracts for the sale of the goods took place outside India, and the goods were also delivered to the buyers outside India. No other operation such as procuring raw materials or manufacture of finished goods took place in India. The Supreme Court had to make its decision keeping in mind the lack of statutory definition of “business connection”. The Court pointed out that in order for a non-resident to have a business connection with India, he had to have a “real and intimate” connection with the business activities occurring in India.3 The court in this case looked into the authority of the representatives while considering how real and intimate their relationship was with the non-resident companies. The representatives had no authority to conclude contracts on behalf of the nonresident companies. The representatives were not acting as agents for the non-resident companies. Their activities led to the making of offers by merchants in India to the non-residents which the non-residents were not required to accept.4 The Supreme Court therefore concluded that there was no “real and intimate” relation so as to

serve as a medium for the courts to engage in an intensely factual inquiry. The law in this area is really all about how the courts weigh the facts of the case. As the courts themselves acknowledge, “no general principles or universal tests” can be laid down and “each case must be determined on the total impression created on the mind of the Court by all facts and circumstances disclosed in that particular case” (para 7). If this is unhelpful, it is not because the courts like to be vague but that the concept of a trade does not allow for precision. 2 1964 Indlaw SC 254. 3 Ibid, para 18. 4 Ibid.

Business Connection

31

establish a business connection. Notice how the lack of authority to conclude contracts rather than the volume of contracts concluded decided the “real and intimate” issue in this case. Aggarwal is significant in that it clearly brought out a qualitative component to the business connection test. The volume of business in India—for example the number of contracts enabled by the representatives—was not enough to establish a business connection in India. The Supreme Court discussed case law that pointed out that there should be an element of continuity between the business of the non-resident and the business activity occurring in India. A stray or isolated transaction is normally not regarded as a business connection.5 To sum up the Aggarwal case, the connection between the business activity of the non-resident and the business activity in India should neither be feeble nor be fleeting.

‘Business Connection’ Post-amendment In the past several years, the government has amended the Act twice to further explain the meaning of “business connection”. The first of these sets of amendments was to align the Indian tax law with the language in double tax avoidance treaties. The amendment focuses on persons in India who are acting on behalf of non-residents. According to this amendment, certain enumerated activities of these persons would lead to the non-resident being considered as earning income through a business connection in India. The amendment provides for three kinds of activities that lead to a business connection in India. It is not entirely clear why the Act and by implication the DTAAs zeroed in on these activities. Moreover the manner in which these three activities have been described can lead to some confusion. First, if a person habitually exercises the authority to contract on behalf of the non-resident (or habitually plays a principal role in the conclusion of contracts by a non resident), a business connection is established. Notice the words “habitual” instead of the more usual phrase “regularly”.6 Habitual connotes a more serious relationship between the representative and the non-resident; since the representative habitually makes contracts on behalf of the non-resident, he is seen as more emphatically (read real and intimate) representing the non-resident in the host country. What kinds of contracts are being envisaged here? On the face of it, the amendment is curiously silent about the kind of contracts it is concerned with, except to describe the contracts in general terms as contracts for the transfer of property and contracts for the provision of services. As can be seen from our discussion above, it is the way in which the representative exercises his authority that seems to concern the amendment, not the substantive content of his exercise of authority. 5 Ibid, para 14, quoting from Abdullabhai Abdul Kadar v. Commissioner of Income-tax Bombay City (1952 22 I.T.R. 241) 1952 Indlaw MUM 2. 6 Section 9(1)(i) Explanation 2(a).

32

4 Permanent Establishment

The second activity of the representative that is considered as establishing a business relationship is if the representative habitually maintains in the host state a stock of goods from which he regularly delivers goods on behalf of the non-resident.7 This provision applies to those non-residents who have some kind of a distribution centre in India from which they deliver goods to customers or to other locations. Again, note the use of the phrase “regularly delivers goods and merchandise”. Long term storage might be excluded under this provision. Like the rest of the amendment there is once again the use of the phrase “habitual” which searches for a more permanent and serious relationship between the representative and the non-resident. How permanent a relationship does habitual denote? Well, it is always been notoriously difficult to identity someone’s persistent activity as his habit and it is perhaps as difficult to pinpoint the habitual selling of the non-resident’s goods as it is to identity an evening constitutional as a habit. The third activity identified in the amendment requires the representative to habitually secure orders for the non-resident in the host country.8 This provision aims at capturing the activities of those representatives who do not enter into contracts on behalf of their principals but rather do pretty much everything else in their authority to secure sales for their principal. The third activity moves the idea of a business connection closer to the facts in Aggarwal. This provision only applies to those representatives who work wholly or almost wholly for their non-resident associates. One intriguing questions that arises here is from whose viewpoint—representative or non-resident—does the concept of “wholly or almost wholly” apply? If it applies from the point of view of the non-resident, then as long as the non-resident has only one agent in the host county, this provision would apply even if the representative also represents many other non-residents in the host country. If it applies from the point of view of the representative, then this provision would apply only if he is acting exclusively in the host country for one non-resident although the same non-resident might have engaged several other representatives in India. How do these rules modify or add to the principles discussed in Aggarwal? If the representative in India has the authority to enter into contracts on behalf of the non-resident, it would lead to a business connection under the amendment. This principle, it can be argued, is already implicit in Aggarwal. But note the phrase “habitually exercises”. What is the reason for this phrase? Perhaps the government— just as the Supreme Court in Aggarwal—wants to emphasise the fact that an isolated transaction cannot come within the ambit of this clause; there must be an element of continuity between the business of the non-resident and the activity in India. Moreover, as mentioned earlier, habitual commercial acts on behalf of someone are a strong indication of a real and intimate commercial relationship. Note the different ways in which the government has used implicitly the concept of “real and intimate” connection in this amendment. One measure of “real and intimate”—in addition to the habitual commercial relationship—can be the extent of the activities performed by the Indian representative for the non-resident principal, 7 Section 8 Section

9(1)(i) Explanation 2(b). 9(1)(i) Explanation 2(c).

‘Business Connection’ Post-amendment

33

and this is reflected in the requirements of the amendment: if the representative has the authority to enter into contracts, or acts as a distribution centre or as a selling agent, his activities can establish a business connection for the non-resident in India. The amendment has another interesting component which we have not addressed yet: the exclusion for independent agents.9 Independent agents of non-residents are excluded from the ambit of the amendment and will therefore not be at risk of creating a business connection for the non-resident in India. The amendment does not define an independent agent but lists illustrations of such agents such as a broker or a general commission agent. The independent agent has to act for his principal in the ordinary course of his business. The independent agent loses his status if he acts mainly or wholly on behalf of his principal. What is the reason for the exception relating to independent agents? One reason could be that the amendment aims to conform to the corresponding tax treaty language (most tax treaties treat independent agents as creating no tax obligations for their principals). Another reason—more in conformity with the principles of business connection in section 9—could be that the Act considers that the independent agents have sufficient autonomy in their decision making process to defeat any notion of a “real and intimate” business connection with their non-resident principals. As is evident, when a non-resident company arranges with an agent to do some business in India, the particulars of the arrangement are very important. The manner in which a company parcels out its work between itself and its agent is a matter of some importance, and the company should take care to ensure as far as possible that either it does not fall within the ITA’s business connection provisions at all or if it does, it is able to utilise the independent agent exception.

Taxability of Business Profits as per DTAAs: Understanding ‘Permanent Establishments’ Where there is a double tax treaty between India and another country, the treaty’s provisions would trump the business connection test. The typical DTAA (in this essay, the UK-India DTAA is used as a model) provides that the business profits of a nonresident company are only taxable in the country of its residence, with the caveat that the company can also be taxed in the country where it has a permanent establishment. I will refer to countries in which a non-resident company does business as host countries. Thus a non-resident company doing business in India (a host country) would be taxed in India if it has a “permanent establishment” in India. A DTAA definition of a permanent establishment generally adopts a three tier structure. First a general definition is provided along with illustrations of the general definition. The intention clearly is to provide for a flexible concept that would apply even if the nature of the business changes globally. However general the definition is, it is still limiting and once we have considered in detail the definition of a permanent 9 Section

9(1)(i) Explanation 2, first proviso.

34

4 Permanent Establishment

establishment, we can appreciate how it has to grapple with the challenges of technology enabled services in contemporary times. Second, the definition provides for exceptions so that certain factors that might be considered as indicative of a permanent establishment are expressly excluded. Third, the issue of agents of non-resident enterprises is dealt with separately.

Fixed PE For example, under article 5 of the UK-India DTAA, the definition in its first tier describes a permanent establishment as a fixed place of business through which the business of the corporate enterprise is carried on. The emphasis on fixed needs some consideration. The concept is looking for some aspect of the foreign corporation that anchors it in the host country. A fleeting presence of a corporation, however lucrative it maybe in terms of profits, does not amount to a permanent establishment. But how is the idea of a fixed presence to be assessed? It is a notoriously difficult question to answer.

Service PE The definition in its second tier provides some illustrations. It provides that the concept of a permanent establishment includes a company’s “anchors” such as a place of management, an office, a factory, a workshop or a sales outlet. But sometime a company is nothing but the sum total of its employees and what if the company simply sends a bunch of employees into India to do some work and they come back after some time. The DTAAs provide for this eventuality as well. Typically, if the employees of a non-resident corporation provide services in India for more than 90 days in a year, the non-resident corporation is considered to have a permanent establishment in India. The number comes down to 30 days if the services are provided to an associated enterprise in India. The first tier of the definition is followed by the second tier of exceptions to the definition. For example, Article 5 of the UK-India DTAA provides that in certain circumstances, even if the circumstances amount to a permanent establishment, they would be deemed to have not fulfilled the definition of a permanent establishment. These circumstances usually relate to the company maintaining a fixed place in India merely for storage, advertisement, display, collection or supply of information or the conduct of scientific research, and for other activities of a preparatory or auxiliary character, thus confirming that only activities that directly result in profits are at risk of being considered as constituting a permanent establishment.

Dependent Agent PE

35

Dependent Agent PE The next level of complication would be the case of an agent PE. If, instead of a fixed location, the foreign enterprise uses an agent, then the foreign enterprise will be deemed to have a PE. Most tax treaties have detailed provisions about the kind of representation that would lead to an agent becoming a permanent agent of the foreign enterprise. Typically three kinds of agent activity would result in a PE of the foreign company. An agent PE is called a dependent agent permanent establishment (DAPE) of the foreign enterprise. First, a person who has and habitually exercises an authority to negotiate and enter into contracts on behalf of the foreign enterprise will constitute the PE of the foreign enterprise unless the contracts are for the purchase of goods and merchandise. Second, a person who habitually maintains a stock of goods from which he delivers goods on behalf of the foreign enterprise would be considered as a PE. Third, a person who habitually secures orders wholly or almost wholly for a foreign enterprise would be considered as a PE of that foreign enterprise. The DAPE is a peculiar form of PE in the sense that a PE is established without the tangible presence of a foreign enterprise in India. The DAPE is a deemed PE that arises solely because of the ability of an India based person to personify the foreign enterprise. The question that arises naturally after the identification of the PE is that of the attribution of profits to the PE for Indian tax purposes. This question is essentially that of a foreign company’s economic nexus with a country. In international affairs, it’s a fundamental maxim that a country cannot tax all of a foreign enterprise’s business profits but only that portion of a foreign enterprise’s profits that have an economic nexus with the country. The idea of an economic nexus is mainly a territorial one. A country ought to have jurisdiction to tax a foreign enterprise only if that foreign company has some physical presence in the country and only to the extent that profits can be attributed to the physical presence. The ITA recognises the economic nexus principle in section 5(2) which states that the income of a non resident is income that arises or is received in India, thus establishing the need for an Indian territorial link between a foreign enterprise and its liability to Indian taxation. Section 9 has laid down more detailed criteria for determining the circumstances in which a particular item of income is deemed to arise or accrue in India. Tax treaties that India has signed with major industrialised countries have also recognised the principle of territorial nexus. When it comes to the taxation of permanent establishments, most tax treaties have a provision that lays out the criteria for the attribution of income to the permanent establishment. A typical provision (for example, Article 7 of the India U.K. DTAA) would state that if a foreign enterprise has a permanent establishment in India, only the profits of the foreign establishment that are attributable to the permanent establishment are taxable in India. Such profits are profits that a permanent establishment would be expected to make if it were a distinct and separate enterprise engaged in the same or similar activity dealing wholly independently with the foreign enterprise of which it is a permanent establishment.

36

4 Permanent Establishment

An attribution provision of this kind transplants the familiar arms length principle into the area of attribution of profits to permanent establishments. The arm’s length principle brings with it associated problems of computation of profits which we will have occasion to see below. Permanent establishments present contracting parties, especially multinational concerns operating in technology enabled services, with multiple opportunities to structure their transactions in a tax efficient manner. Through adroit contracting, they can make use of Indian human resources and claim not to have a fixed business presence in India. A typical situation involves a parent in a developed country jurisdiction (assume it is a UK company) and a fully owned Indian subsidiary. From a contractual point of view, these are two separate entities, each with their own articles of association setting out their respective management structures. Traditional contract law respects the independent commercial reality of the parent and the subsidiary. Any contractual dealings inter se between the parent and the subsidiary should not lead ipso facto to any adverse consequences so far as the parent’s risk of having a permanent establishment in India is concerned. The OECD Model as well as the India-UK DTAA recognises the independent legal status of the subsidiary. Article 5(6) of the UKIndia DTAA states that a subsidiary in the host country does not by that very reason become a permanent establishment of its parent company.

Stewardship v. Deputation A parent can contractually choose to contract with it subsidiary in the following manner. It can outsource complex technological functions to its Indian subsidiary, for which it would pay the subsidiary an arm’s length price. In addition, it might enter into another arrangement with its Indian subsidiary whereby it deputises some of its employees to India to direct the work of the Indian subsidiary. The salaries of the deputed employees would be paid by the parent company but the subsidiary would reimburse the parent company for an amount that would be equivalent to the employee compensation. The contractual picture presented is one of two independent legal entities dealing with each other at arm’s-length. However the Indian revenue have never liked these arrangements and have always attempted to establish a tax nexus over the UK parent on the basis of its relationship with its subsidiary. According to the Indian revenue, the arm’s length price paid by the parent to the subsidiary is not the right amount to tax. The right amount to tax is the profits made by the parent by using the subsidiary’s services. That this diminishes the independent legal status of the parent and the subsidiary as well as the contractual relationship between the parent and the subsidiary is a by product, in the revenue’ view, of the rightful imposition of the permanent establishment status on the subsidiary. The facts described above were in play in the Supreme Court case of DIT (Mumbai) v.

Stewardship v. Deputation

37

Morgan Stanley.10 In Morgan Stanley, the employees deputed by the UK parent to the Indian subsidiary were categorised into distinct groups. One group, the deputationists, worked at the Indian subsidiary presumably to steer the operations of the Indian subsidiary in the right direction. The second group, the stewards, worked at the Indian subsidiary for quality control purposes, in order to ensure that the Indian subsidiary had the systems in place that were appropriate to handle confidential client information and allied matters. The Supreme Court made a distinction between the two kinds of employees. It said that the deputationists alone would be at risk of triggering a service PE. The stewards, since they did not provide any services for others but instead were merely working to safeguard the interests of their employer, were not engaging in an activity deserving of a PE status. The deputationists did spend a considerable time in India and their activities would result in a service PE.11 After characterising the deputationists work as resulting in a service PE, the Supreme Court curiously termed the Indian subsidiary to be the PE of the American parent.12 Normally, if an American company, through its employees, performed services in India beyond the time limits prescribed in the India-US DTAA, it would be the American company that would be considered as having a PE in India as a result of its employees work. However, in Morgan Stanley, the Supreme Court decided that the deputationists’ work for the Indian subsidiary resulted in the Indian subsidiary being the PE of its American parent company.13 One way in which the characterisation of the Indian subsidiary as a service PE can be explained is to hold that the steering work done by the deputationists resulted in the Indian enterprise (that they were steering) the service PE of the foreign enterprise. When the work done by the foreign enterprise’s employees is of a managerial character, then it is not the work done by the employees alone that constitutes the service PE; the Indian subsidiary steered by the deputationists itself becomes the service PE. This transformation of a service PE from a bunch of employees working in India to the Indian subsidiary steered by the deputationists can have knock down implications on the profits that can be attributed to the service PE, a subject that will be revisited below. In judicial decisions subsequent to Morgan Stanley, the imposition of status over contract comes across in stronger terms. In Centrica India Offshore Pvt. Ltd v. CIT,14 the Indian subsidiary, Centrica, was formed by a UK parent to monitor the quality of back office operations performed for the UK parent by third party Indian service providers. The UK parent seconded some of its employees to Centrica to assist Centrica with its monitoring and quality assurance activities in the initial stages of Centrica’s operations. The employees remained on the payroll of the UK parent.

10 DIT

Mumbai v. Morgan Stanley, 2007 Indlaw SC 739. paras 13 and 14. 12 Id para 14. 13 See VanderWolk (2007) where the author points out this anomaly with the Supreme Court decision. 14 Centrica India Offshore Pvt Ltd v. CIT, (2014) 364 ITR 336. 11 Id

38

4 Permanent Establishment

However, Centrica had agreed to reimburse the UK parent to the extent of the salaries earned by the seconded employees during their time in India. The Delhi High Court gave a decision which undermines the contractual structures in place between the parent and the subsidiary in two significant ways. First, it classified the money spent by the subsidiary on the seconded employees as fees for technical services (FTS) and not as a reimbursement for expenses.15 Second, the court also took the view that the UK employees’ work in India would result in a service PE for the UK parent. Centrica is unusual as it not only made the UK parent taxable for what was characterised by the entities as mere reimbursements for costs (of sending employees to India) but also recognised the possibility of additional taxation because of the UK parent’s service PE in India. The most recent decision on facts similar to that found in Morgan Stanley and Centrica is the Supreme Court decision in ADIT v E Funds IT Solutions Inc.16 The Indian subsidiary of an American parent company was performing back end support services for the parent company. The parent company deputed its employees to the subsidiary company for a certain period of time in return for which the subsidiary reimbursed the parent company. The Supreme Court did not recognise an Indian PE for the American parent company. In an important development over previous law, the Supreme Court clarified that in certain cases the nature of services performed by the employees of the parent company might be such that it does not amount to a business activity in the host state. As long as the subsidiary’s activities were not customer facing, the Supreme Court appears to consider such activities as only ancillary in nature and therefore not deserving of the permanent establishment status.17

‘Nature of Business’ and ‘Control’ Test Another important Supreme Court decision on permanent establishments relates to the Formula One World Championship races conducted in India.18 The Formula One car racing business is founded on a set of ‘commercial rights’ held by a US company, called the Formula One World Championship Limited (FOWC) which in turn exploits these rights by appointing persons in various countries around the world to host, stage and promote the Formula One race under a Race Promotion Contract (RPC). The RPC enumerates a certain amount of consideration to be paid by each country’s entity to FOWC. In India, a company called Jaypee Sports International Ltd (Jaycee) entered into a RPC with FOWC, and paid FOWC a certain sum of money

15 Id. 16 (2018)

13 SCC 294. para 26: “As has been noticed by us … 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.”. 18 Formula One World Championship Ltd. v CIT, International Taxation, (2017) 15 SCC 602. 17 Ibid,

‘Nature of Business’ and ‘Control’ Test

39

for the right to host, stage and promote the Formula One race in India. The race was to be carried out at a site called the Buddh International Circuit. The Indian revenue argued that the Buddh International Circuit was a permanent establishment of FOWC in India and thus asserted its jurisdiction to tax the income of FOWC that was attributable to India. The question before the Supreme Court was whether the Buddh Circuit fulfilled the permanent establishment criteria under the India-United States double tax treaty, since FOWC was a US company. The relevant test under Article 5 of the treaty was whether the foreign enterprise had a fixed place in India through which the business of the foreign enterprise was carried on. The Supreme Court stated that there was no doubt about the fact that the Buddh Circuit was a fixed place in India.19 Since the Buddh Circuit was used to conduct races (which generated profits), it was equally evident that some business activity was being carried out through the Buddh circuit.20 Since the Buddh Circuit was a fixed place of business in India and some business activity was being carried out through it, the key remaining question was who had control over the Buddh Circuit at the time the business activity, namely the racing, was taking place? The question of control was characterised by the Supreme Court in the following manner: Was the Buddh Circuit at the disposal of FOWC? If so, FOWC had a permanent establishment in India. The court undertook to answer this question by looking at the whole gamut of contractual arrangements between FOWC and Jaypee, making it clear that the court was not interested in looking at any particular contract in isolation. While the agreement between FOWC and Jaypee gave Jaypee the right to host, manage and promote Formula One racing in India, Jaypee transferred the media and title sponsorships rights, as well as paddock rights, to affiliates of FOWC. The affiliates in turn transferred the media rights to an Indian entity. Under a separate agreement with a third party, FOWC undertakes to provide licensing and supervision services at the Indian event as well as services related to travel, transport and date support needed for the Indian event. Therefore it was clear that despite the RPC, FOWC and its associates continued to exercise control over the Indian event. The RPC, read on its own, also gave FOWC considerable control over the Indian event. Clause 5(e) of the RPC stated that the race circuit would be prepared with the approval of FOWC and FIA.21 The RPC obliged Jaypee to ensure that the Circuit was open to FOWC, its affiliates, its licensees and its contractors during the race, fourteen days before the race and seven days after the race.22 With respect to the areas of the Circuit not open to the public, Jaypee could not grant access except through passes issued by FOWC.23 Jaypee was prevented under the express terms 19 Ibid,

para 75. para 75, where the Supreme Court stated ‘From this circuit different races, including the Grand Prix is conducted, which is undoubtedly an economic/business activity.’. 21 FIA is the ‘Federation Internationale de I’ Automobile’, a regulatory body that makes rules and regulations for Formula One races. 22 Para 77, analysing clause 11 of the RPC. 23 Id, noting clause 14 of the RPC. 20 Ibid,

40

4 Permanent Establishment

of the RPC to make any audio-visual recording within the Circuit.24 Further, Jaypee had to ensure that personnel accredited by FOWC were allowed inside the circuit to make audio-visual recordings and provide all such help and facilities that FOWC would require.25 While it was clear from the aforesaid provisions that the Buddh Circuit was under the control or at the disposal of FOWC, the Supreme Court had to consider whether the limited time period of the race in India meant that the Buddh Circuit could not in any case be considered as a fixed place of business because of the inherently transient nature of the racing business. Much like the High Court on the issue, the Supreme Court emphatically rejected the argument that the limited time period of the Buddh circuits operations derogated from the fixed nature of the Buddh Circuit. The Supreme Court first took note of the High Court’s point that the nature of the business would impact the fixed place of business analysis.26 The Formula One racing was sporadic by its very nature. Further, the RPC was for a five year-duration and during the six weeks that FOCW had access to the Circuit, its access was exclusive. Given these factors, the High Court was of the opinion that FOWC had a fixed place of business in India even if the race itself, which was the fount of its profits, was conducted in a very short period of time. The Supreme Court also referred to several cases to buttress its point. One of the cases referred to by the court was Joseph Fowler v M.N.R.27 This case involved a travelling US salesman who used to spend three weeks every year in Canada at a fair using a camper trailer to sell his products. The US salesman was considered to have a permanent establishment in Canada even though his duration of stay in Canada was for a very limited period of time, as the idea of permanent establishment takes its cues from the nature of the business in which it is implicated. Having decided that the Buddh Circuit was a fixed place of business and was at the disposal of FOWC, the court discussed the third element of its inquiry into the existence of a permanent establishment: did FOWC conduct a business in India? The permanent establishment test looks for three indicators, a place in the host country that can qualify as a durable place of business, a certain amount of control or the right to disposal to the place of business in the hands of a foreign enterprise, and finally a business (rather than a sporadic set of activities) conducted by the foreign enterprise through the fixed place of business. At the third stage of the permanent establishment inquiry, it had been established that the business of the foreign enterprise in the host country was not fleeting. The only remaining question was whether the foreign enterprise’s economic activities in the host country were feeble enough that it escaped the appellation of a permanent establishment. The inquiry into the third stage is into the mechanics of the foreign corporation’s economic activities in India. The Supreme Court took note of the High Court’s observations in this respect: while Jaypee owned the Buddh Circuit, its contribution to 24 Id,

noting clause 18.1 of the RPC. noting clause 20.1 of the RPC. 26 Para 79. 27 (1990) 90 DTC 1834: (1990) 2 CTC 2351. 25 Id,

‘Nature of Business’ and ‘Control’ Test

41

the business conducted through the circuit was only a part of a larger business empire which was completely controlled by FOWC. The important drivers of the business were the choice of the circuits where the races would be held, the participation of the racing teams and the media rights associated with the races. In all these matters, it was FOWC and not Jaypee that played the leading role.28 The Supreme Court had no hesitation in holding that FOWC’s economic activities in India were far from feeble and that it did conduct a business in India through its permanent establishment. It’s interesting to note that both the Supreme Court and the High Court arrived at their decisions though a close analysis of the various contracts that were in play, especially the so called Concorde Agreement between FOWC and the racing teams and the RPC agreement between FOWC and the host country promoters, thus underlying the importance of contractual terms to the permanent establishment characterisation.

Attribution of Profits The inquiry into the existence of a permanent establishment does not exhaust the taxability question for a foreign enterprise in India. As discussed above, the next question after the identification of a permanent establishment relates to the attribution of profits (for Indian tax purposes) to the permanent establishment. The Indian case law on attribution of profits to foreign enterprises because of their permanent establishments paints a complex picture that is exacerbated by the operation of foreign subsidiaries in India. In the simplest case, a foreign company has a permanent establishment in India because it has a branch office or a factory in India, which constitutes its location PE. In this case, it should be fairly easy for the revenue to ascribe profits to the foreign company’s PE since it would be the profits earned by the branch or the location PE, for which accounts would be available. The same analysis can be made for installation projects. However, the attribution of profits in the case of DAPE is especially complicated because it is not immediately apparent how to attribute profits beyond looking at the arms length price paid to the dependent agent by the foreign company. Current case law indicates that there is no need to look beyond the arms length price but the matter is not free from doubt. In DDIT, International Taxation v Set Satellite (Singapore) (Pte.) Ltd.,29 the ITAT had a different view of how income of the foreign company ought to be taxed when the foreign company had a DAPE in India. The ITAT made a distinction between the income of a dependent agent PE (which conforms to an arms-length income) and the income attributable to a foreign company as a result of deemed PE by virtue of the DAPE.30 In the hypothetical example given by the ITAT, a Singapore company

28 See

para 82 and the text from the High Court judgement quoted therein.

29 MANU/IU/5227/2007. 30 Ibid,

para 10.

42

4 Permanent Establishment

sources electronic equipment worldwide and sells the equipment in India through a DAPE.31 The DAPE earns a commission on the Indian sales and incurs expenditure in arranging the sales. If, in attributing income to the foreign company, there is no need to go beyond the arms length commission paid to the DAPE, then the foreign company’s Indian tax liability is exhausted as long as the commission earned by the DAPE is arms length commission. The ITAT did not believe that this is the appropriate manner in which Indian source income must be attributable to the Indian DAPE.32 According to the ITAT, the starting point of the income attribution ought to be the total value of the sales to Indian customers conducted through the DAPE. From this figure, the costs incurred by the foreign company should be deducted. Such costs would comprise the purchase price of the products sold to Indian customers, the handling charges of the foreign company, the sales commission paid to the DAPE and any reimbursements of the DAPE costs by the foreign company.33 The final figure might well exceed, by a considerable margin, the arms length commission paid to the Indian DAPE. The ITAT’s approach towards PE income attribution in the case of a DAPE is known as a double entity approach, where in addition to the dependent agent in India, another deemed entity namely the deemed permanent agency is created as a result of the dependent agent. However, the Bombay High Court overruled the ITAT decision on the point of attribution, as a result of the Morgan Stanley decision.34 Matters are even more complicated when the foreign company has an Indian subsidiary. The Indian subsidiary might, through its employees, enter into contracts on behalf of the foreign company. If it does so, it becomes a dependent agent PE and then the same question arises regarding attribution of profits: should the attribution go beyond the profits earned by the subsidiary? Alternatively the Indian subsidiary might host deputationists from the parent company, in the context of helping the Indian company performing outsourcing activities for the parent company. In such a case, as per the approach taken by the Supreme Court in Morgan Stanley, the Indian company will become the foreign company’s service provider in India (because of the steering role played by the deputationists) and then the question is whether the arms length compensation paid to the subsidiary would be sufficient attribution of profits, with due credit given to the payment made by the subsidiary to the parent. This raises the same issue as the one identified in the dependent agent cases. The Supreme Court in the Morgan Stanley case stated that there was no need to go beyond the arms length price. Finally, it must be noted that the relevant DTAA will have an impact that not only on the calculation of the income of the PE but also on the calculation of the expenses that the PE can deduct for tax purposes. For example, in DDIT v. Unocol Bharat 31 Ibid,

para 10. paras 15, 16 ad 17. 33 Ibid, paras 22 (discussing the OECD model report) and 24. 34 SET Satellite v DDIT, 307 ITR 205 (2008), in particular paras 12 and 13, where the Bombay High Court discussed the Supreme Court judgment in Morgan Stanley. 32 Ibid,

Attribution of Profits

43

Ltd,35 the assessing officer denied the expensing of certain payments made by a PE because the PE did not withhold tax on those payments. The restriction on expensing based on the non withholding of taxes on payment is found in section 40(a)(i) of the ITA. The ITAT pointed out that the relevant DTAA (India-Mauritius) allowed for the deduction of expenses by the PE without any additional restriction found under relevant domestic law and therefore section 40(a)(i) will not have any effect in the context of the India-Mauritius DTAA.

Conclusion The case law discussed in this chapter has shown that to seek clarity with respect to the meaning of permanent establishments and the attribution of income to permanent establishments, one must begin with the basic building blocks of a permanent establishment found in the OECD model tax convention. The OECD model has been followed in its basic structure by most of the industrialised nations. The PE provision unfolds in five progressive stages. The first stage begins with a general statement of permanent residency that defines a PE as a fixed placed of business through which the business of an enterprise is carried on. The second stage identifies some common forms of the PE like an office or a factory. The third stage identifies situations where a PE does not exist such as activities that are merely preparatory or auxiliary in nature. The fourth stage identifies dependent agents as permanent establishments. The fifth stage makes it clear that a subsidiary, merely because of its subsidiary status, will not be a permanent establishment of its parent. Once a PE is established, the OECD model tax convention has a separate article that attributes income to the PE on arms length independent entry basis. The most striking feature of the permanent establishment definition at the first stage (a fixed place through which the business of the enterprise is carried on) is the confluence of time, space and substance. The idea of a fixed place connotes a space within the host country which is at the disposal of the enterprise and a certain minimum time (depending on the business of the PE) for which the enterprise makes use of the space at its disposal. The requirement that the activity carried on through the permanent establishment has to be the business of the enterprise shows the importance of the nature of the activity carried out within the time and the space occupied by the enterprise. The requirement is that the activity has to have some substance from the business point of view. An ancillary or marginal commercial activity (whatever might be ambiguities involved in characterising something as marginal or ancillary) will not do. Therefore at the first stage, the basic theoretical basis of a PE has been set out, and the implications of time, space and substance can be noticed in the difference between an actual establishment and a virtual establishment. In the case of an actual establishment, the foreign enterprise has a presence in the host state that combines the three elements of space, time and substance mentioned 35 MANU/ID/0899/2018.

44

4 Permanent Establishment

above. This would be the case when the foreign enterprise has a factory or an office in India. In the case of a virtual establishment, the three elements of space, time and substance are fulfilled by a person other than the foreign enterprise but all three are then attributed to the foreign enterprise. This would be the case where the foreign enterprise has a dependant agent in the host country. The defendant agent will operate an office and even though the foreign establishment does not have physical access to the office, the spatial, time and substance elements of the office will be attributed to the foreign enterprise because of the relationship between the foreign enterprise and the dependent agent. The discussion above makes it clear that in addition to an actual permanent establishment, the PE provisions in the OECD model convention also envisage a virtual permanent establishment formed through the DAPE provisions discussed above. The DAPE PE is virtual because the foreign enterprise does not actually have a physical establishment in the host country but instead has some other person’s physical establishment attributed to it. While the PE of this kind is virtual, it is important to remember that there is a reality behind it-of space, time and substance, except the space, time and substance is attributed rather than actual. In addition, the DTAAs entered into by India envisage another kind of virtual establishment, which is established by the provision of services in the host country by the employees of the foreign enterprise. For example, the India-U.K. DTAA has a provision in the second stage PE definition (the stage that identifies specific instances of PE) which provides that if employees of a foreign enterprise spend more than 90 days in any twelve month period in the host country, the foreign enterprise will have a PE in the host country.36 The number of days required to form a PE is reduced to 30 in the case of services performed by the employees of a foreign enterprise for an associated enterprise in the host country.37 Much like the DAPE virtue, the service PE is also a virtual PE because the space, time and substance established by the employees in the host country are attributed to the foreign establishment of which they are employees. The reason for such attribution is that the employees are one kind of dependent agents of the foreign enterprise, the other kind being the more familiar ones who work on behalf of the foreign enterprise for a commission. Let us call the PE formed by the employees of the foreign enterprise as an employee based DAPE rather than the commission based DAPE. The constitution of an employee based DAPE is slightly more complicated than that of a commission based DAPE. In the case of the commission based DAPE, the requirements of space, time and substance are easier to find, as the commission agent is likely to work out of an office, whether temporary and permanent, to which he has the right of disposal. In the case of employees, while time and substance should not be an issue, finding a space to which they have a right of disposal is more problematic. It is easier to find if they work out of a hotel or out of their client’s offices to which 36 Article 5(2)(k)(i). The article mentions ‘employee and other personnel’ but for the present, this chapter will not analyse the meaning and implications of ‘other personnel’ and will instead focus on the central case of employees. 37 Article 5(2)(k)(ii).

Conclusion

45

they have a temporary right of disposal. But what if the employees of the foreign enterprise are seconded to its Indian subsidiary? The usual arrangement for such a secondment is for the Indian subsidiary to provide a space to the seconded employees but this space is within the control and at the disposal of the Indian subsidiary. The seconded employees work for and under the control of the Indian subsidiary as de facto though not de jure employees of the Indian subsidiary. The seconded employees steer the work of the Indian subsidiary. For the secondment of the employees, the Indian subsidiary makes a payment to its foreign parent, which reimburses the foreign payment for the salary it pays to these employees. It is in these situations that I believe the Supreme Court in Morgan Stanley was constrained to treat the Indian subsidiary itself as the service PE of the foreign enterprise. It is the subsidiary who has the right to disposal over the space in which the requisite time and substance needed to establish the PE are being achieved. The time element is satisfied by the presence of the employees for the required minimum number of days in the host country. The substance element is being satisfied by the work that the subsidiary performs under the guidance of the seconded employees. The subsidiary, because it provides the space, controls the employees, has its work steered by the employees, and is the subsidiary of the foreign enterprise, becomes a de facto employee DAPE of its parent. The complexity of identifying a virtual PE is mirrored in the problem of attributing income to a virtual PE, once a virtual PE is identified. The ITAT in Set Satellite adopted the double entity approach and asked the following question: what income of the foreign establishment (out of its global income) has to be attributed to the virtual establishment in India, which consists of the space, time and substance belonging to the dependent agent. This question is subtly different from the question asked with respect to actual establishments: what would be the income earned by the actual establishment if it were to be considered as an independent entity dealing with the foreign enterprise (of which it is a part) on an arm’s length basis. The latter question is interested on the income earning activity of the actual PE in order to attribute income to the actual PE whereas the former question is interested in the income earning activity of the foreign enterprise that it derives through its virtual Indian PE. Both approaches seek attribution to the Indian PE but the perspectives of the two approaches are different. The actual establishment attribution looks at the Indian PE’s contribution to the global income of the foreign enterprise whereas the virtual establishment attribution looks at the foreign enterprise’s global income that can be sourced to the Indian PE. The Supreme Court in Morgan Stanley does not appear to have appreciated the difference between an actual PE and a virtual PE when it came to the question of attribution of income. After identifying the Indian subsidiary as the PE of its foreign parent (which has been described as a virtual PE in our discussion), the Supreme Court was content to look at the question of the attribution of income as a question of attribution for an actual PE. This meant that the question was whether the income earned by the Indian subsidiary for the services that it performed for its parent was an arm’s length income. The Supreme Court continued to seek an arm’s length income

46

4 Permanent Establishment

for the subsidiary PE whereas it should have discussed an arm’s length income for the virtual PE formed because of the activities of the subsidiary. The arms length income for the virtual PE will be the income earned by the foreign enterprise because of the activities of the deputed employees in the space provided by the subsidiary. This income could be equal to or more than the money paid to the subsidiary. If it is equal to the money paid to the subsidiary, then no further income will be taxed in India as the net income of the virtual subsidiary will include the compensation to the subsidiary. If it is more, the difference will have to be taxed in India as the income of the virtual PE. One further complications need to be note here. If the business conducted in India by the subsidiary is only back end services for its parent, then as per E Funds there is no PE issue because the substance aspect of the PE model is not met. The activities conducted by the subsidiary simply do not rise to the level of business that is required to constitute a PE.

Reference VanderWolk J (2007) News analysis: Indian Supreme Court’s decision in Morgan Stanley Poses a PE Puzzle. WTD 157-7

Chapter 5

Royalty and Fees for Technical Services

Introduction Status has always played a strong role in the taxation of royalties and technical service fees. Even though payments for royalty and technical services would normally come within the remit of business income, most of the tax treaties signed by India provide for a separate tax regime for royalties and technical services. The India-UK DTAA is one such example. The provision on business income gives way to the provisions on royalty and technical services.1 Therefore, when a UK company earns India sourced income from royalty or technical services, it cannot make use of the permanent establishment exception to shield its income from Indian taxation. Once its business income has acquired the quality of royalty or technical services, its earnings are no longer classified as business income under the India-UK DTAA. India has similar but more stringent domestic tax provisions regarding royalty and technical services. Normally, income from royalty and technical services would come under the ambit of business income under the ITA. However, in 1976, the ITA was amended to introduce section 115A, which imposes a special rate of tax on income from royalty and technical services earned by non-resident companies. Section 115A does not define what amounts to royalty or fee from technical services, instead it defers to section 9 of the ITA on these issues, which is the section that decides what is deemed to be Indian source income for Indian tax purposes. Therefore it transpires that the provisions in section 9 relevant to royalties and technical services determine the source of these payments as well as, confusingly, define royalty and fees for technical services. The royalty and technical service provisions in section 9 are examples of the Indian government’s effort to impose status over contract, by widening the wedge already 1 Article

7 (7) of the India UK DTAA states that the provisions on business income will not apply in the case of income from royalties and technical services (termed as fees for included services), which are addressed separately in Article 12.

© The Author(s), under exclusive licence to Springer Nature India Private Limited 2020 N. Nuggehalli, International Taxation, SpringerBriefs in Law, https://doi.org/10.1007/978-81-322-3670-2_5

47

48

5 Royalty and Fees for Technical Services

provided in section 115A. Once it is granted that whatever might be the contractual relationships amongst parties, it is the statutory provisions that count, the stage is set for the statutory provisions to be expanded further to suit the requirements of the revenue. The discussion below will outline the sourcing and definitional provisions relating to royalty payments and FTS in section 9 followed by an analysis of the recent controversial amendments to these provisions. According to the source provisions in section 9, royalty payments and FTS would be deemed to have an Indian source if the royalty payments and FTS were made by an Indian resident.2 This provision is fairly uncontroversial and has not given rise to much litigation. If on the other hand, the royalty payments and FTS are made by a non-resident, the payments would nevertheless have an Indian source if the payments were made to service a business or profession carried on in India or to earn income from a source in India.3 This latter deeming provision is more controversial as it could potentially apply to royalty payments or FTS made by one non-resident to another non-resident, thus extending Indian jurisdiction over business transactions between two non-residents. However, this provision has been in place for nearly forty years and does require, at the very least, some Indian connection, however tenuous. As shall be discussed below, the Indian government has made the Indian connection even more tenuous through a recent amendment.

Royalty-Definitional Provisions The definitional provisions in section 9 contain, much like the sourcing provisions, another set of deeming provisions. The original definition of royalty in section 9 was meant to capture four kinds of payments with respect to intellectual property. One kind of payment was concerned with payments for obtaining a right to trade with respect to intellectual property. Examples of this kind are payments for the right to lease patents, trademarks, and copyright obtained from the owners of such patents, trademarks and copyrights.4 For example, an Indian company contracts with an American company to obtain the rights to its software so that it could resell or lease the software to its Indian customers. Another kind is concerned with payments for the right to possess and control the intellectual property, without obtaining the right to trade in it. Examples of the second kind are payments for the commercial use of patents, trademarks, copyright and equipment.5 An Indian company involved in financial information processing obtains a license from a UK software company to use UK software to build a financial software, which would involve access to the base code of the UK company’s software. The third kind is concerned with payments for 2 Section

9(vi)(b) and (vii)(b). 9(1)(vi)(b). 4 Explanation 2(i) to section 9(1)(vi). 5 Explanation 2(iii) to section 9(1)(vi). 3 Section

Royalty-Definitional Provisions

49

obtaining knowledge pertaining to intellectual property, without obtaining either the right to trade with respect to intellectual property or the right to possess and control the intellectual property.6 Examples of the third kind are payments for the imparting of technical or scientific information. The fourth kind is perhaps the simplest instance of royalty payment, which is the payment for the right to use the product of intellectual property, obtaining neither the right to trade, nor the right to control, nor any access to information with respect to intellectual property. Examples of the fourth kind are payments for straightforward leases of equipment. The equipment is a product of technology; however the lease of technology does not involve either a transfer of technical knowledge (at least not necessarily) or a transfer of a right to trade in the intellectual property that was used to produce the equipment or the transfer of a right of control and possession of the intellectual property that was used to produce the equipment. The inspiration for these provisions are the royalty provisions found in the model tax treaties of the OECD, UN and US versions, and incorporated in various DTAAs between countries. The OECD model, which was the first to address the problem of defining royalties, defines royalties as the consideration for the use of and the right to use intangible property.7 The UN and US Models follow a similar approach to the definition of royalty, although there are some differences between the three models. The definition of royalty, both in the model DTAAs as well as in the Indian tax provisions, does not make a distinction between different kinds of intangible property. As a result, when one wants to distinguish between royalty income on the one hand and business income on the other, one has to consider the ideas of use and right to use. When a person spends money for the use or the right to use intangible property, his payments are characterised as royalty. Such payments are different from payments made to acquire intangible property, not merely to use or obtain the right to use property. The use of rights terminology can be traced to the idea that property is nothing but a bundle of different rights, and that a transfer of ownership of a certain property involves a transfer of the entire bundle of rights associated with such property. One doesn’t transfer ownership in property if one has retained some of the rights in the bundle and transferred the rest to the transferee. Consequently the distinctness between transfer of rights to use property and transfer of ownership in property hinges on whether the transferor has retained some rights while transferring some other rights to the transferee. If he has, the payments he received are royalty. If he hasn’t, the payments he receives comprise either business income or capital gain.

6 Explanation

2(ii) to section 9(1)(vi).

7 Article 12. The article describes intangible property as copyright, patent, trademark, design, model

plan, secret formula or process and information concerning industrial, commercial or scientific experience.

50

5 Royalty and Fees for Technical Services

Royalties and the Transfer of Software The problem with the distinctions mentioned above is these distinctions can become quite difficult to apply when most of the rights associated with intellectual property are transferred. Because some rights, however trivial, have been retained, one can argue that the payment received for the transfer of the rights is the payment for the use of or for the right to use intangible property. On the other hand, it is somewhat stretching the language of transfer of rights to apply this language to a situation where the transferor retains only a simulacrum of rights, where all the important rights that enable a person to use and enjoy intangible property have been given away. In the business of transfer of rights, there are two additional factors that complicate the issue of whether the payment received for the transfer of rights is royalty or business income (I will ignore for the time being the issue of capital gain). Both these complications are found in the sale of off the shelf or packaged software to Indian residents from abroad. First, packaged software is sold through discs or similar physical packages and the customer, who is either an Indian end-user or, more frequently, an Indian distributor of software products, pays a sum of money to acquire the software package. Further, while the Indian customer or distributor does not obtain any right to exploit the software, usually, each package provides a limited licence to each enduser customer granting the customer the right to use the software on the customer’s computer. The language of licence is misleading; the customer is unable to process the software any further: all he can do and is legally allowed to do is to use the software on his computer. Sometimes, the end user customer also obtains a right to make a back up copy of the software on his computer. The Indian High courts have not decided uniformly on this matter, and the Supreme Court of India has not yet taken cognisance of this matter.8 The Karnataka High Court in Commissioner of Income Tax v Samsung Electronics Co Ltd.9 had to decide whether payments for shrink wrap software (off the shelf software) ought to be characterised as royalty or business income. The stakes were high as the seller of the software did not have a permanent establishment in India. Therefore, the characterisation of the payment as business income would enable the seller to escape taxation altogether and excuse the buyer from any withholding tax obligations. The Karnataka High Court approached the issue of royalty characterisation by way of an analysis of the Indian Copyright Act, 1957, which provides copyright protection to literary works. The court looked into the licensing agreement between the non-resident and the Indian based purchasers to understand the nature of the 8 In

Tata Consultancy Services v State of Andhra Pradesh, 271 ITR 401, the Supreme Court ruled that off the shelf software was ‘goods’, which might then provide credence to the idea that the transfer of such software for consideration is akin to the sale of goods. However, Tata Consultancy was decided in the context of a legislation on sales tax. By its very nature, this decision cannot decide whether, for purposes of the income tax legislation, payment for software rights would be characterised as royalty or business income. 9 2011 SCC OnLine Kar 3973 (Hereafter Samsung).

Royalty-Definitional Provisions

51

commercial relationship established between them as a result of the transfer of the software. The agreement for the transfer of software to the end users provided for an individual, non-transferable and non-exclusive license to use the software on the terms and conditions mentioned in the agreement.10 The agreement with the distributors was of the similar kind except that the distributors were granted an exclusive licence in their territories. All agreements, whether with distributors or the end users followed a similar patterns of grant of limited rights with respect to the software. The users of the software agreed that all right, title and interest in the software remained with the non-resident transferor and that the transferee could not resell the software. Further, the users could not copy (other than to the extent permitted) or modify any part of the software or associated documentation. Given the limited nature of the rights transferred under the agreement, the court came to the conclusion that there was only a transfer of a copyright and not a transfer of ownership of software.11 The revenue, in arguing for royalty treatment for the transfer of software, had one hurdle to cross. Curiously this hurdle had been created by another revenue department, this time the sales tax department of the state of Andhra Pradesh (AP). The AP sales tax department had successfully argued before the Indian Supreme Court that canned software, because it was sold in packaged form, was in the nature of tangible movable property and therefore taxable as the sale of goods under the Andhra Pradesh (AP) sales tax legislation.12 The upshot of the Supreme Court judgment was that shrink wrapped software had, by virtue of the medium through which the software was being transferred, alchemised into tangible property from being intangible property in its original state. The buyers of software in the Samsung case argued that because shrink wrapped software was considered as goods under the sales tax legislation, a transfer of such goods would be outside the purview of royalty provisions under the ITA. The customer buying the shrink wrapped software is no longer paying money for the copyright in the software; he is instead paying money for the physical product in which the software is housed. No longer is the subject of the commercial transaction a limited right to an intellectual property, the subject of the transfer is now a physical good like any other tangible commodity such as a book or a television set. In other words, the medium of the transfer of software determines the nature of the product transferred. In response, the Karnataka High Court pointed out that the Tata Consultancy case was decided under a different tax legislation.13 Even if the shrink wrapped software are classified as goods under a sales tax legislation, that did not in any way preclude 10 Samsung,

para 32. para 32, where the court stated that ‘the contention of the learned senior counsel appearing for the respondents (the buyers of shrink wrapped software) that there is no transfer of copyright or any part thereof under the agreements entered into by the respondents with the non-resident supplier of software cannot be accepted.’. 12 Tata Consultancy Services v State of Andhra Pradesh, (2004) ITR 401. The question of the classification of canned software as goods arose under the Andhra Pradesh General Sales Tax Act, 1957. 13 Samsung, para 34. 11 Ibid,

52

5 Royalty and Fees for Technical Services

the possibility of the transfer of shrink wrapped software giving rise to royalty under an income tax legislation. What would be considered as a royalty payment under the income tax legislation is a matter exclusively for income tax jurisprudence. Therefore the taxpayer’s reliance on Tata Consultancy was disfavoured because of its irrelevancy to the income tax issue at hand. The Samsung decision relied solely on whether the property transfer that was the subject of the dispute in the case was the kind of property that was protected under the Copyright Act, 1957 (CA). If yes, payments for the transfer of such a property would be considered as royalty under the ITA. Therefore the definition of copyright under the CA was key to the classification issue in Samsung. The CA defined a copyright as the right to reproduce a literary, dramatic or musical work or the right to copy such work or perform such works. The CA recognises computer programmes as a literary work and protects the rights (such as the right to copy) associated with computer programmes. Since the right to copy a computer programme is a right recognised under the CA, it would follow that the transfer of shrink wrapped software would result in royalties as the software provides a right to the purchaser to make a copy of the software for his purposes. However, there is a complication introduced by the CA itself in this regard. Section 52 (aa) of the CA provides that if the lawful possessor of the software makes a copy of the software either in order to use the software for the purposes for which it was supplied or in order to make a back up copy, such acts would not constitute an infringement of the copyright in the software. The Samsung decision acknowledged this provision but made it clear that the presence of this provision did not affect the copyright issue at all. On the contrary, the court pointed out that the exception was only available to a lawful possessor of copyright, and therefore anyone who used the exception was by definition entitled (and would have paid for) copyright to the software.14 The court’s interpretation of the effect of section 52 of the CA is rather unfortunate as a more natural interpretation was available to it. The court could have interpreted section 52 of the CA to mean that even persons who do not have a copyright in the software but have instead purchased a physical medium (of which the software is a part) are allowed, as lawful possessors, to use the software on their computers or to make a back up copy of the software. This interpretation is more natural as it quite reasonably allows the copying or use of software in typical commercial situations where the recipient of the software does not have a copyright to the software. The Samsung decision assumed mistakenly that one needs to have a copyright to software in order to make use of the exception mentioned in section 52 of the CA. Samsung is important not only for its role in interpreting the meaning of royalty for domestic tax purposes. Most treaties with India do not define royalties; instead the treaties allow domestic law to determine the meaning of royalty. Therefore, the Samsung decision has the effect of denying treaty benefits to taxpayers as well, since 14 Samsung,

para 37, where the court points out that it is because of the licence granted that the licencee is able to claim that making a copy of the software or using the software on the computer does not give rise to an infringement of the copyright in the computer.

Royalty-Definitional Provisions

53

they are now precluded from claiming that a more tax payer friendly meaning is available in the relevant treaty. In the case of Samsung, the treaties in question were those with the United States of America, France and Sweden, but the same result can be seen in the case of treaties with similar provisions such as for example the treaty signed with the United Kingdom. Indian courts have not been unanimous on the issue of the classification of software. The Delhi High court took a view that was different from the one expressed in Samsung. In Director of Income Tax v Infrasoft Ltd 15 (hereafter Infrasoft), there was a supply of customised software from a non-resident company to customers in India. The software was customised to suit the requirements of customers and was licensed to the customers after the customisation. The Indian branch of the non-resident followed up the supply of the software with after sales installation and customer training. The assessing officer classified the payment for the software as a royalty payment based on a number of reasons. It was argued that the payment was for providing a right to copy the software (on to the computer systems of the customers) which was a kind of right that was protected under the CA and therefore a royalty payment. In the alternative, it was argued that since the software was the result of sophisticated scientific research, the customer had made a payment for the imparting of scientific information, and was thus a royalty payment as recognised in section 9 of the ITA. Finally, it was argued that since the price of the software was related to the commercially valuable (and secretive) function performed by the software, the customer actually paid a price for the use of a secret formula or process, a category of payment recognised as royalty under section 9 of the ITA.16 The arguments of the revenue showcased the malleable nature of the definition of royalty under the ITA and the possibility of the same transaction being captured under different provisions of the ITA definition relating to royalty. The Delhi High Court, much like the Karnataka High Court, approached the classification issue from the point of view of the Copyright Act, 1957 (CA); however it came to a different conclusion. The court pointed out that the CA, in section 14 enumerated the rights accorded to the copyright holder. These rights, as mentioned earlier, relate to the exploitation of the intellectual property (IP) in the copyright such as the right to reproduce the copyrighted article in some form or the other, the right to issue copies of the copyrighted product to the public, the right to stage a performance in public that is based on the IP (if the IP lends itself to a performance), and the commercial renting of the products of the IP (a common occurrence in the case of computer programmes where the software is leased out to a customer in the form of a product). It followed therefore that if the purchaser of software did not acquire any of the rights held by the copyright holder, the purchaser cannot be said to have made royalty payments to the copyright holder. The Delhi High Court took a certain overarching approach towards the nature of copyright under the CA that informed its understanding of what it means to 15 Director

of Income Tax vs. Infrasoft Ltd., 2013 SCC OnLine Del 4694. para 11.

16 Infrasoft,

54

5 Royalty and Fees for Technical Services

earn money by exploiting one’s copyright. The court emphasised the exclusivity dimension and the public dimension of the rights recognised under section 14 and stated that none of the rights recognised under section 14 were available to the clients of the supplier of software.17 On the face of it, it is perhaps difficult to understand why one’s exploitation of one’s copyright must end in providing exclusive rights to the transferee of the copyright. Perhaps the idea is that the right to enjoy property, say a house, necessarily requires an exclusive right to use the house. If a person is paying a sum of money to rent a house, then in the normal course of things, the right to live in the house for a duration of time is an exclusive right as at the same time other people are not given the same right to live there. The supposition here is that the concept of a transfer of a copyright (which is what gives rise to royalties) has an idea of exclusivity built into it. Similarly, the Delhi High court emphasised the public dimension of a copyright as, presumably, a transfer of a copyright is of no effect if the transferee does not have a right to exploit the right in relation to the wider public, either by reproducing the object of the copyright for sale to members of the public or by performing the object of the copyright in the public (in the case of a play, for example). Therefore, in the perspective of the Delhi High Court, the exclusivity and public dimensions are intrinsic to the concept of a copyright, and would colour any interpretation of the CA.18 The Delhi High Court went through the contract documentation related to the transfer of the software and concluded that all ‘copyright and intellectual property rights’ remained with Infrasoft.19 This conclusion is technically incorrect, as the transferee did acquire a limited right to reproduce the software for making archival or back-up copies. The court regarded this right as of little consequence and in fact, one of the cases cited in Infrasoft described the outcome of such a right as “plainly contrary” to the provisions of section 14.20 In fact the right is not plainly contrary to the rights recognised under section 14 unless one believes that a copyright necessarily involves the right to publicly (and therefore commercially) exploit the copyright. As pointed out earlier, the Karnataka High Court took a contrary stand and regarded any right, even if unconnected to its commercial exploitation, to be a right the transfer of which can give rise to a royalty payment. The Delhi High Court also considered section 52(1)(aa) of the CA which provides that making a back up copy of the software to guard against loss or destruction would not be considered as an infringement of copyright. The Court quoted the Authority for Advance Ruling decision in Dassault Systems KK, IN Re where the Authority considered the presence of section 52(1)(aa) to mean that such rights (right to make backup copies) are not to be considered as rights protected by the mechanism of a 17 Infrasoft,

para 87. is recognised explicitly in the language of the CA. Section 14 of The CA begins with the words, ‘For the purposes of this Act, “copyright” means the exclusive right, subject to the provisions of the Act…’. 19 Infrasoft, para 86. 20 Infrasoft, para 65 citing Motorola Inc v. Deputy CIT, para 159. 18 Exclusivity

Royalty-Definitional Provisions

55

copyright as recognised in section 14, CA.21 It is interesting that a consideration of the same section led the Karnataka High Court to take the opposite view i.e. that the presence of section 52 (1)(aa) (which applies only to a lawful possessor of a copy of a computer programme) means that, ordinarily, (without the section 52 safe harbour) making a back up copy without an express license to do so would amount to a breach of a copyright held by the transferee. The nature of a copyright does not undergo a change only because one instance of its breach is legislatively excepted. In fact, according to the Karnataka High Court, since section 52 (aa) only applies to lawful possessors of copyright, this is a conclusive indication that the transferee of software must already have a copyright in the software in order for the legislature to further empower such copyright with certain additional rights. Therefore any person granted the right to make backup copies must in exchange be paying royalties for certain rights, including the legislative right to make backup copies. The flaws in this argument have been noticed earlier in the analysis of this decision. The judicial dispute over the classification of software is not settled yet, and it is likely to be the subject matter of a decision by the Indian Supreme Court. In the meantime, the Indian government attempted to settle this issue statutorily, as we shall see below.

Retrospective Amendments to the Statutory Definition of Royalty Despite the quite broad based sourcing and definitional provisions in section 9, the Indian government continues to be concerned with technology driven international income for two reasons. First, it is concerned that modern technology, particularly digital technology, makes it inevitable that technology might be transferred for consideration to Indian customers through contractual mechanisms that would ensure that India would not be considered as the source state for such income under traditional tax rules. This can be termed as the problem of nexus erosion. Second, it is concerned that non-resident companies might fashion their contracts with Indian customers to ensure that what might traditionally result in royalty income would instead be classified as business income, which is not taxable in India without a permanent establishment. This can be termed as the problem of classification erosion. In the last five years, the Indian government amended section 9 provisions on royalties and technical services to address the twin concerns of nexus erosion and classification erosion. The problem of nexus erosion has been done away with in exemplary fashion. Section 9 was amended to make it clear that income from royalties and technical services would be deemed as arising in India even if (a) the person who has earned royalty income or fees for technical devices does not have a place of

21 Infrasoft,

para 74 citing Dassault Systems KL, IN RE (2010) 322 ITR 125 (AAR).

56

5 Royalty and Fees for Technical Services

business in India and (b) has not rendered any services in India.22 What this means in practice is that an Indian payer and a UK payee can have a perfectly sanitized cross border contractual relationship in which there is a transfer of technology or technological knowledge without the UK payee having any physical presence in India whatsoever (much less a permanent establishment), and yet section 9 will deem such income to have an Indian source solely on the basis of the fact that the payor is an Indian resident. The problem of classification erosion posed different problems with respect to royalty and technical services respectively. The government was concerned that nonresidents will try to classify royalty received from Indian residents as straight sales of goods, and they were particularly concerned about the transfer of off-the-shelf software (e.g. sale of software via discs). As discussed previously, when it comes to transfer of software through discs, many software vendors provide a limited license to the purchasers to make a copy of the software on the customer computers or make a back-up copy to meet emergency outages. No rights of modification, further sale, etc. are conferred by these licenses. The vendors took the position that the mode of transfer (limited licensing) and the method of transfer (in physical disc form) were a complete answer to the issue of whether the income earned in the off the shelf software sales comprised royalty income or plain vanilla business income. The vendors contended that the income earned was business income. In 2012, the high water mark of the triumph of status over contracts, the government decided to foreclose the issue by adding two explanations to the definition of royalty in section 9. The first explanation closed out any contractual mechanism that tried to address the classification issue through the sale of discs by stating that computer software refers to any computer programme recorded on any disc, tape, perforated media or other information storage device.23 The second explanation closed out any contractual mechanism that tried to address the classification issue through limited licensing by stating that the ‘transfer of rights’ in the definition of royalty includes any right to use a computer software including granting of a license.24 The government introduced further amendments in section 9 to counter the Delhi High Court decision in Asia Satellite Telecommunications Co. Ltd. v. DIT.25 A nonresident television broadcaster had paid a sum of money to a non resident satellite company to use their transponders to transmit signals to Indian cable providers and television viewers. The revenue argued that the payments made by the television broadcaster were royalty chargeable under the ITA. The court disagreed and gave two 22 Explanation 5 to section 9(1)(vi) states: “For the removal of doubts, it is hereby declared that for the purposes of this section, income of a non-resident shall be deemed to accrue or arise in India under clause (v) or clause (vi) or clause (vii) of sub-section. (1) and shall be included in the total income of the non-resident, whether or not,— (i) the non-resident has a residence or place of business or business connection in India; or (ii) the non-resident has rendered services in India.”. 23 Explanation 3 to section 9(1)(vi). 24 Explanation 4 to section 9(1)(vi). 25 2011 SCC OnLine Del 507.

Retrospective Amendments to the Statutory Definition of Royalty

57

independent reasons for its decision. First, royalties could arise only if the transponders were under the control and possession of the payor.26 Further the source of the payments was outside India. The payment was made by a non-resident to another non-resident. The payment was for a service that was ultimately directed towards Indian customers but this was considered too remote to convert the payment into an Indian source payment.27 The government, in response added Explanation 5 and 6 to section 9(1)(vi). Explanation 5 made it clear that certain characteristics of payments were irrelevant when it came to identifying such payments as royalty payments. It is irrelevant whether such payments resulted in the payer possessing or controlling any right, property or information. Similarly, it is irrelevant whether the payer uses the right, property or information directly. Finally it is irrelevant whether the location of such rights, information or property is in India or otherwise. In sum, the third explanation did away with the Asia Satellite decision’s link between royalty on the one hand and possession, control, use and Indian situs on the other hand. Explanation 6 went one step further. It clarified that the expression ‘process’ includes transmission by satellite, cable, optic fiber or any other similar technology, whether or not such process is secret. This provision once again aims at neutralizing Asia Satellite by ensuring that the satellite transmission charges are definitely considered as a technological process, the payment for which is considered as royalty. Despite the government’s efforts at attacking legislatively the problem of classification erosion, the revenue continues to be suspicious of taxpayer attempts at classification erosion through allegedly ingenious contractual arrangements. The most recent example of revenue’s attack on contractual arrangements that the revenue perceived to be a device of classification erosion can be seen in the M/s Google India Private Ltd. v. Addl. Commissioner of Income-tax (hereafter Google AdWords) case decided by the ITAT Bangalore bench.28 The Google Group, headquartered in the Unites States had, among other subsidiaries, two subsidiaries each in Ireland and India respectively. Google Ireland and Google India had a two level commercial relationship. At one level, Google India performed typical information technology enabled services (ITES), which were back end services meant to ensure that Google Ireland met its client’s technical requirements and was in compliance with legal regulations. The services provided by Google India for Google Ireland involved the use of Google software and algorithms. These services were performed pursuant to an ITES contract that laid down the respective obligations of Google India and Google Ireland. At another level, Google India also acted as a distributor for Google Ireland in reselling online advertising space offered by Google Ireland. The reselling was pursuant to a distributorship agreement between Google India and Google Ireland.

26 Ibid,

para 68. paras 72 and 73. 28 2017 SCC OnLine ITAT 78. 27 Ibid,

58

5 Royalty and Fees for Technical Services

Therefore there were two different contractual relationships between the same corporate entities. One was with respect to providing some back end technical services and the other was with respect to reselling advertising space offered by Google Ireland. The revenue was concerned that the Google group, through its contractual mechanisms, was not paying adequate tax on the amount of money it was earning though advertisements from Indian customers. A large amount of money was flowing from Google India to Google Ireland as the purchase price paid by Google India for the online advertisement space offered by Google Ireland, which Google India resold to Indian customers. Under the contractual arrangements as envisaged by Google, the Indian revenue was able only to tax Google India’s profits margins on its re-sales to Indian customers. The money flowing to Ireland was taxable only in Ireland, which because of Ireland’s low tax regime was a tax bonanza for Google. The revenue decided to characterise payments from Google India to Google Ireland as royalty payments, although, on a perusal of the ITAT judgment, the precise grounds on which the revenue was prepared to arrive at its royalty characterisation was not clear. Nevertheless the royalty characterisation was sufficient for the revenue to assert its right to impose a withholding tax obligation on Google India for its payments to Google Ireland. The ITAT made a detailed survey of the use of Google technology and came to the conclusion that advertisements placed by customers through Google could not be equated to advertisements placed by customers in newspapers and magazines.29 When a newspaper charges for an advertisement, they are merely selling a physical space in which the customer is allowed to place a pre-determined text provided by the customer. An advertisement placed through Google is different from a newspaper advertisement because of several reasons. The advertisement is not a mere selling of space but involves the application of technology by Google that makes the advertisement appear online as a function of key word searches and customer browsing histories.30 The Google Adwords Programme used technology to enable the customer to reach its market. The customers paid for the use of technology (mainly software developed by Google) and not for the sale of advertising space. Does this use of technology determine the characterisation of the customer payment as royalty? The ITAT answered this question in the affirmative.31 The definition of royalty has been considered in this chapter and Google demonstrates yet again the complexities involved in the royalty characterisation. It is remarkable how definitional issues continue to play a pivotal role in tax disputes to this day. The point of contention continues to be the question of what kinds of rights have to be given away such that a transfer of a bundle of rights will result in business income rather than royalty income. The sale of space in a newspaper is a transfer of a bundle 29 Ibid, para 38 onwards. See in particular, para 52, where the Tribunal states: ‘On the basis of above, in our view the agreement between the assessee and the Google Ireland was not in the nature of providing the space for advertisement and display the advertisement to the consumers.’. 30 Ibid, paras 53–55. 31 Ibid, para 60.

Retrospective Amendments to the Statutory Definition of Royalty

59

of all rights relating to a particular space in a newspaper on a particular day. However, the same characterisation is difficult to make regarding the Google-customer relationship. Google does not sell advertisement space in the straight forward way that a newspaper sells it. Instead it interacts with its customer needs using its software so as to offer a customised service that is both flexible (the customer can change the key words to which the advertisements respond) and dynamic (the software, in conjunction with the customer, chooses the internet users to which the advertisements can be targeted, and the target population can change over a period of time). Given the above characterisation of Google’s business, the question is whether Google India obtained either contractual rights or proprietary rights from Google Ireland. Without obtaining proprietary rights, one finds it difficult to understand how there can be any question of royalty characterisation. A royalty characterisation would require a partial proprietary right to be provided to the rights recipient. The ITAT drew a line through the business model of Google right through to the relationship between Google Ireland and Google India. However I believe further justification is needed before one goes from the Google business model to royalty characterisation. The fact that Google ads are rich with use of services using intellectual property does not mean that the middleman or the distributor of such intellectual property services is also exploiting such intellectual property. The ITAT Google Adwords judgement demonstrates the need to understand better the basic nature of the rights the acquisition of which lead to royalty treatment. Perhaps one can argue that the question of exploiting or using property, which has always been a key element of the definition of royalty, is receding into insignificance because of the insertion of Explanation 5 to the definition of royalty, according to which it is irrelevant whether the person allegedly paying royalty needs to be in control or possession of the relevant intellectual property. Nevertheless I believe that the definition of royalty, especially when it applies to cases of intellectual property being used for some purposes, continues to demand that some right of exploitation or use (what I have termed as proprietary rights) to have been transferred between parties. A judicial resolution of the ambit of royalty under section 9 after the amendments is awaited.

Fees for Technical Services-Definitional Provisions Much like the provisions concerning the definition of royalty, the ITA in section 9 contains a broad definition of FTS. Explanation 2 to section 9(1)(vii) states provides that FTS refers to any consideration for the rendering of any managerial, technical or consultancy services. Therefore a wide swathe of activities is covered under the definition of FTS and it is not the case that FTS is confined only to what might be considered as technology rich services. The position under tax treaties is slightly different. To begin with, the US and the OECD model treaties do not address the issue of FTS specifically. Neither model treaty has a specific provision relating to FTS. In the event a DTAA follows any of these models, consideration under FTS would be

60

5 Royalty and Fees for Technical Services

business income and a non-resident in the host country would escape taxation unless the non-resident has a permanent establishment in the host country. Recently, the UN model treaty included a provision addressing fees for technical services.32 In most of the treaties entered into by India with other countries, India has succeeded in addressing specifically the issue of the taxation of technical services. India wants to tax such services in much the same fashion as royalty payments, which would mean that a non-resident offering technical service with an Indian connection would be liable to taxation in India even if that non-resident does not have a permanent establishment in India. A typical example of this approach is Article 13 of the UK-India DTAA, which entitles the host state to tax payments for technical services according to the law of the host state provided the technical services ‘arise in’ the host state. Although section 13 gives primacy to the law of the host state, it defines the term ‘technical services’, which therefore serves as a limitation on the taxing powers of the host state to the extent the definition of technical services in the Article is more restrictive than the corresponding definition in the domestic law of the host state.33 According to article 13(4), technical services are of two kinds. Either these services are ancillary to the transfer of intellectual property or scientific equipment that generates royalties or these services are technical services in their own right. It is the latter that give to interpretational problems. Technical services in the latter category are defined in article 13(4) as services that make available technical knowledge, experience, skill, know-how or processes or consists of the development and transfer of a technical plan or technical design. Notice how the India-UK DTAA definition of technical services differs from the domestic statutory definition in significant respects. It requires that the technical knowledge, experience or skills are made available to the transferee of the technical services. A typical case that brings out the significance of this phrase is the Delhi Income Tax Appellate Tribunal decision in Income Tax Officer, TDS v Nokia India Pvt Ltd.34 The assessee, an Indian company, hired a contractor to design and set up a manufacturing concern in India. However, in order to ensure that the contractor met quality control standards, the assessee consulted a Finnish company to review the manufacturing and construction plans of the contractor. The Indian revenue argued that the payments received by the Finnish company were fees for technical services and therefore the assessee was under an obligation to withhold taxes on the payments made to the Finnish company. The India-Finland DTAA had provisions on the definition of technical services that were pari materia with the India-UK DTAA.35 Therefore the question of whether 32 Article

12A, UN Model Double Taxation Convention (2017 update). this with article 14 on capital gains, which, subject to certain exceptions mentioned therein, allows capital gains to be taxed according to the law of the respective contracting party. Since Article 14 does not define capital gains, the definition of capital gains is also left to be decided under the law of the respective contracting party. 34 2015 SCC OnLine ITAT 9865. 35 The make available specification in the definition of technical services is a standard clause in many of India’s tax treaties including the India-US DTAA. 33 Contrast

Fees for Technical Services-Definitional Provisions

61

any technical services were made available to the assessee was the key to the tax issue in this case. The tribunal stated that in order for technical services to fall within the ambit of the DTAA, it was essential that technological knowledge or skill (considered broadly) was transferred to the recipient such that the recipient was enabled to use the technology by himself without depending further on the transferor.36 The review services performed by the Finnish company did not transfer any technological knowledge or skill that enabled the assessee to use this knowledge in its business. The Finnish company’s remit was to review the designs, diagrams, costestimates and other parameters of construction that would enable it to assess the construction on the basis of the quality standards required by the assessee.37 The Finnish company provided a service but not a service that resulted in transferring knowledge or skill in the sense described above. The tribunal therefore held that the services performed by the Finnish company could not be technical services within the meaning of the India-Finland DTAA. Despite the narrow definition under the DTAA clauses, in other respects the Indian government has cast its net wide when it comes to taxing technical services. Once again, it is the territorial nexus issue that has most bedevilled the issues in this area. The problem of technical services and territorial nexus began with the Supreme Court decision in Ishikawajma-Harima Heavy Industries Ltd. vs. Director Of Income Tax, Mumbai.38 According to section 9(vii) of the ITA, consideration for technical services would be deemed to arise in India if (a) the consideration is paid by the Government or (b) an Indian resident unless the Indian resident is utilising the technical services in a business or profession carried out outside India or to earn income from a source outside India or (c) the consideration is paid by a non-resident and the technical services are utilised by the non-resident in a business or profession carried out in India or to earn income from a source in India. Ishikawajma involved a turn-key contract to build an Liquefied Natural Gas (LNG) facility in India for which payments were being made by an Indian company, Petronet LNG Ltd. In connection with this project, the assessee performed some technical services offshore which were distinct from the rest of the contract. The Supreme Court had to consider whether technical services performed offshore by the nonresident but utilised in India were taxable in India. The Supreme Court considered section 9 (vii)(c) and stated the following: Reading the provision in its plain sense, it can be seen that it requires two conditions have to be met; the services which are the source of the income that is sought to be taxed, has to be rendered in India, as well as utilized in India, to be taxable in India. In the present case, both these conditions have not been satisfied simultaneously, therefore excluding this income from the ambit of taxation in India. Thus, for a non-resident to be taxed on income for services, such a service needs to be rendered within India, and has to be a part of a business or profession carried on by such person in India. The Petitioners in the present case have

36 Nokia

India, para 9.

37 Id. 38 (2007)

3 SCC 481.

62

5 Royalty and Fees for Technical Services provided services to persons resident in India, and though the same have been used here, it has not been rendered in India.39

It is surprising that the Supreme Court relied on section 9(vii)(c) since that subclause applies only if the payments for services flow from a non-resident. In this case, the payor of the consideration was Petronet, a resident. Consequently, it would be section 9 (vii)(b) that would apply. Section 9 (vii)(b) has a very different set of rules. According to section 9 (vii)(b), consideration paid by a resident for the use of technical services would normally be considered as Indian source income unless the services are utilised by the payor (i.e. the Indian resident) in a business carried out by the resident outside India or the services are used by the resident to earn income from a source outside India. If the Supreme Court had held that section 9(vii)(b) was applicable, there would have been no doubt that the assessee was taxable as the consideration received by the assessee was paid by an Indian resident (Petronet) and was utilised in the Indian resident’s business in India. Nevertheless the Supreme Court’s comments on section 9 (vii)(c) have become canonical and have been widely cited by courts and tax scholars. Section 9 (vii) (c) taxes the consideration for services that is paid by a non-resident but only if the services are utilised in a business carried on by such person in India or the services are used to earn income from any source in India. One reasonable interpretation of this provision is that the non-resident receiving the services must either utilise these services in a business in India, thus establishing a territorial connection to India or these services must be used to earn income from any source in India thus establishing a financial connection to India. An example of a territorial connection to India would be the case where a French company is paid by an American company for technical services that would be utilised in the American company’s business in India.40 An example of a financial connection to India would be the case where a French company is paid by an American company for technical services that would be utilised by the American company in providing a service or product to customers resident in India even though the American company does not carry out any business in India. Indeed, in cases of a financial connection to India, as long as the American company is earning money from Indian customers, it is not necessary that the American company have any physical footprints in India. If one adopts this interpretation, the question of territorial nexus, while important, is only one of the ways in which the Indian revenue can assert their right to tax consideration for technical services received from a non-resident. They have, in addition, the option to tax using the financial nexus that has been provided for in section 9(vii)(c). However the Supreme Court in Ishikawajma did not adopt this interpretation. Applying the concept of a territorial nexus, the Supreme Court held that the technical services fees received by the non-resident would be taxable only if the assessee’s 39 Id

para 90. the business carried out in India by the American company would result in a permanent establishment under the ITA or for the purposes of a tax treaty is an additional tax issue which is not considered here. 40 Whether

Fees for Technical Services-Definitional Provisions

63

services were utilised as well as rendered in India. Since the assessee had not rendered the technical services in India, the technical service fees did not have an Indian source.41 The government responded to the Ishikawajma decision by introducing an explanation to section 9(1)(vii) in 2007. According to this explanation, which applied retrospectively from 1976, any technical services income was considered as Indian source income under section 9(vii) regardless of whether the non-resident (rendering the services) had a residence or a place of business or business connection in India. Subsequent decisions on the sourcing rule for technical services demonstrated the inadequacy of this amendment to counter the Ishikawajma decision. The 2007 amendment had not made any reference to the relevance of the services being rendered in India. Consequently, the revenue was faced with unfavourable decisions that continued to hold that as long as the technical services were not rendered in India, the fees for such technical services would not have an Indian source. In 2010, the government modified the 2007 explanation to section 9 and made it clear that for purpose of the sourcing rules for income from technical services (among other activities), it did not matter whether the service provider had a residence or a place of business in India, and it did not matter whether the service provider had rendered his services in India.42 The 2010 amendment is yet another example of the triumph of status over status, brought about through legislative fiat. However, as pointed out above, it does stand to reason that section 9(vii) does not conceive of territorial nexus as the sole factor of tax assertion, preferring instead to also include financial nexus as an alternate connecting factor for tax purposes. The doctrine of financial nexus is satisfied when the ultimate consumers of the business enriched by the technical services are Indian residents. Despite the broad language used in section 9(1)(vii), there are cases in which the revenue has failed to prove the existence of a fee for technical services. These cases deal with situations where there is payment involved for the provision of services, but the payment is not considered as consideration for technical services.43

41 Ishikawajma,

para 90.

42 The explanatory memorandum to the Finance Bill 2010 referenced the Karnataka High Court deci-

sion in Jindal Thermal Power Company Ltd. vs DCIT (TDS), which had opined that the Ishikawajma ratio will survive the 2007 amendments to the following extent: A non resident would have to render his technical services in India for such services to be subject to Indian taxation. 43 Deputy Commissioner of Income Tax (International Taxation) v. Welspun Corporation Limited, 2017 SCC OnLine ITAT 334.

Chapter 6

Capital Gains

The Indian income tax schedular system includes, as one of its heads of income, a tax on capital gains. Section 45 charges an income tax on any profits or gains arising from the transfer of a capital asset. A capital asset has been defined widely as property of any kind.1 An Indian resident is taxed on capital gains worldwide whereas a nonresident would be taxed only on capital gains that arise in India. Consequently, we are back to the rules of section 9, which specify the circumstances under which a capital gain is said to arise in India. Until recently, the rules of section 9 on the source of capital gains were quite liberal. When it came to capital gains, the sourcing rules in section 9 focused on the situs of the capital asset that was the subject matter of the transfer. A capital gain that arose, whether directly or indirectly, through the transfer of a capital asset situate (sic) in India was deemed to have arisen in India.2 A provision of this sort was ripe for the plucking; non-resident companies took advantage of this provision when (effectively) transferring the shares of Indian companies owned by them. The best example is the notorious Vodafone case.3 In Vodafone, the vendor took advantage of the inherent ambiguity in the section 9 sourcing provisions regarding capital gains. Shares are capital assets and are situated in India if these shares belong to a company that is incorporated in India. This much is uncontroversial and is accepted both by the revenue and the taxpayers. Therefore shares of India incorporated companies (Indian companies) are situated in India. Shares of companies incorporated outside India (non resident companies) are situated outside India. What this meant was that by virtue of section 9, if a non The part of this chapter on the post Vodafone amendments has appeared in Nigam Nuggehalli, ‘India’s Implementation of the BEPS Project: A Critical Survey’ in Parthasarathy Shome (ed), Insights into Evolving Issues of Taxation (CCH 2016) Part 6. 1 Section

2(45), ITA. 9(1)(i). 3 Vodafone International Holdings B.V. v. Union of India and Anr., (2012) 1 SCR 573. 2 Section

© The Author(s), under exclusive licence to Springer Nature India Private Limited 2020 N. Nuggehalli, International Taxation, SpringerBriefs in Law, https://doi.org/10.1007/978-81-322-3670-2_6

65

66

6 Capital Gains

resident transferred the shares of a non resident company, the capital gains would not arise in India and consequently would be outside the jurisdiction of the Indian tax authorities. What was interesting with this legal regime was that it did not look through the non-resident company to see what it owned. The assets of a non resident company could consist only of shares in Indian companies and yet section 9, at least on a natural reading, would continue to treat the transfer of the shares of the non-resident company as outside the Indian government’s taxing jurisdiction. This lacuna in the application of section 9 was in issue in Vodafone. Vodafone consisted of the most tax efficient manner of affecting the transfer of control over an Indian company. In Vodafone, both the buyer and the seller were non-resident companies. Simplifying the facts, the seller transferred control over an Indian company to the seller by transferring the shares of a non-resident company that owned the share capital of the Indian company. The taxpayer argued successfully in the Supreme Court that section 9 did not have a provision to see through the assets of a non-resident company. The Supreme considered the Vodafone transfer as outside the power of the Indian tax authorities. In order to assert tax jurisdiction over a Vodafone like transfer, the revenue argued for a more capacious model of statutory interpretation, an interpretation that disallowed taxpayers from taking advantage of the provisions of tax statutes in order to minimise their tax liabilities. This kind of argument was a throwback to Chinnappa Reddy’s judgement in McDowells, which had been rejected in the ABA judgement. The revenue once again asked the Supreme Court to reconsider ABA, and argued that ABA, a two judge bench, in effect blunted the effect of the constitution bench in McDowells. The Supreme Court rejected the revenue’s request to reconsider ABA. The court went through the majority judgements and the separate judgment of Chinnappa Reddy. It stated that there was no inconsistency between ABA and McDowell; both judgements were against ‘artificial and colourable’ devices only, and not against legitimate tax avoidance.4 The Indian government’s response was a series of simultaneous amendments to the ITA that radically altered the conceptual foundations of capital gains in Indian tax law and extended the ability of the Indian government to tax cross border capital gains. As many as four amendments were made to the ITA in order to counter indirect transfers of Indian property. These amendments together comprise the triumph of status over contract in the area of capital gains. First, the meaning of a transfer of a capital asset was modified. The definition of a transfer is pivotal to the taxation of capital gains as it marks a stage without which the capital gains taxable event cannot occur. The ITA does not define a transfer exhaustively; instead it has a provision that classifies a number of instances as transfers including straightforward instances such as the sale or exchange of assets.5 The government added an explanation to this provision that included several different means of deriving value from an asset. The transfer of a capital asset now includes a 4 Id

para 64. 2(47), ITA.

5 Section

6 Capital Gains

67

disposal of an asset and a parting with an asset or a disposal of or parting with any interest in an asset. Further a transfer also includes creating any interest in any asset. The explanation also makes sure that any kind of disposal of or parting with or creation of an interest in an asset is captured as a transfer for the purposes of the ITA. In order to ensure this, the explanation says events such as a disposal, a parting or a creation can occur in any manner whatsoever, directly or indirectly, absolutely or conditionally, voluntarily or involuntarily, by way of an agreement or otherwise. Finally the explanation specifically addresses the Vodafone issue. One way to look at the Vodafone case is that while the buyer and seller were interested in transferring control over an Indian company, they instead achieved the same result by transferring the shares of a non resident company that owned the Indian company. In other words, there were two transfers of which one transfer (the transfer of the right to manage the Indian company) was effected by or dependent on or flowed from another transfer (of the shares of a non-resident company). The last part of the explanation addresses this situation, by including, within the meaning of a transfer, a transfer of rights (such as the right to control an Indian company) even if such a transfer is effected by or dependent upon or flowing from the transfer of shares of a non-resident company. Second, the government added an explanation to the definition of a capital asset.6 A capital asset, as explained above, has been defined to mean property of any kind. The government’s explanation clarified that the reference to property carried within its ambit any rights in or in relation to an Indian company, including rights of management or control of any other rights. This explanation was once again specifically aimed at Vodafone like situations. The third and fourth amendments were to section 9 of the ITA. As explained earlier, section 9, which is concerned with income that is deemed to arise in India, has different deeming rules for the different heads of income. For income arising from capital gains, section 9 provides that all income arising through the transfer of a capital asset situate in India would be deemed to arise in India. The third amendment added an explanation that clarified the meaning of ‘through’ and the fourth amendment added an explanation that clarified the meaning of ‘situate in India.’ The third amendment clarified that the expression ‘through’ shall mean and include by means of, in consequence of or by reason of. The expansion of the meaning of ‘through’ makes sense from the point of view of what the government was trying to do, even if can be argued that the explanation does violence to the natural meaning of through. If the transfer of a non-resident holding company’s shares is the means through which the shares of an Indian company are to be (indirectly) transferred or if the transfer of a non-resident company’s shares is in consequence of the (indirect) transfer of the shares of an Indian company, such a transfer would now be deemed to arise in India. Similarly, in an indirect transfer, if the reason shares in the non-resident holding company are being transferred is because the shares in an Indian company are to be (indirectly) transferred, such a transfer would now be deemed to arise in India. 6A

capital asset is defined in section 2(14) of the ITA.

68

6 Capital Gains

The fourth amendment added an explanation clarifying the circumstances in which the shares of a non-resident company would be deemed to be situated in India, which would therefore bring the transfer of such shares within the Indian tax jurisdiction. According to this explanation, the shares of a non-resident company would be considered as situated in India if the non-resident company’s shares derive, directly or indirectly, their value substantially from assets located in India. The 2015 Finance Act has clarified by way of an amendment to section 9 that the share of a non-resident company shall be deemed to derive its value substantially from the assets (whether tangible or intangible) located in India, if the value of Indian assets, (a) exceeds the amount of ten crore rupees; and (b) represents at least fifty per cent of the value of all the assets owned by the non-resident company. India’s ability to impose status over contract in international tax law has run into some difficulties because India has also signed several bilateral investment treaties with the major economies that impose certain obligations of fair treatment on the Indian government in its treatment of investors from contracting countries. A typical example is the India-Netherlands bilateral investment treaty (hereafter BIT). The India-Netherlands BIT includes three kinds of protection accorded to investors. Investors are assured of ‘fair and equitable’ treatment in the territory of the contracting parties.7 Further, investors are assured that they would be able to repatriate their earnings and any gains from the transfer of their property in the contracting states.8 Finally, investors are also assured that their property will not be expropriated by the governments of the contracting states.9 Following the retrospective amendments, the Indian government once again pressed a fresh tax demand on Vodafone on the strength of the newly amended section 9. This time, Vodafone began arbitration proceedings under the IndiaNetherlands BIT (as the purchaser in the Vodafone-Hutchinson deal was a Netherlands entity in the Vodafone group). Among other claims, Vodafone is alleging an absence of a fair and equitable treatment because of the operation of the retrospective legislation.10 The India Netherlands BIT contains a provision that states that the fair and equitable treatment guarantee will not apply to the operation of tax legislation.11 This provision will be a major impediment in Vodafone’s efforts in asserting its claim under the BIT. Vodafone is not alone in being subject to the vagaries of retrospective tax legislation. Recently the Cairns group, headquartered in the United Kingdom, was hit with a multi-billion dollar tax demand on an internal reorganisation consisting of a share transfer between two non-resident entities in the group. The share transfer related to 7 Agreement

between the Republic of India and the Kingdom of the Netherlands for the promotion and protection of investments, Article 4. 8 Ibid, Article 7. 9 Ibid, Article 5. 10 Vodafone seeks a ruling that the Indian government is in ‘breach of its obligations under the BIT, including its obligations under Articles 4(1), 4(2), 4(5) and 5(1) of the BIT’, as quoted in Vodafone’s arbitration claims reproduced in associated litigation before the Delhi High Court. See Union of India v Vodafone Group PLC United Kingdom and Anr, 2017 SCC OnLine Del 9930, para 25. 11 India-Netherlands BIT, Article 4(4).

6 Capital Gains

69

a company that had underlying Indian assets. Cairns brought arbitration proceedings under the India-UK BIT since the tax demand was made on a UK entity within the Cairns group.12 The advent of BIT related dispute resolution has put a new twist on the Indian government’s ability to impose status over contact in tax matters. Any legislation cannot wish away the taxpayer’s contractual relationships by fiat. The taxpayers, if their jurisdiction has signed a BIT with India, now can assert their contractual autonomy in two circumstances. They can resist a legislation that imposes a tax by changing nexus rules retrospectively, and they can also challenge legislation that discriminate between Indian companies and foreign companies in the imposition of tax rules. The Indian government has responded to the use of BITs by aggrieved companies by introducing a model BIT containing provisions that differ from the language in the existing BITs. The model BIT’s language is significant because several BITs signed by India are up for renewal and India is bound to negotiate on the basis of the new language. The model BIT language effectively takes tax matters out of the remit of BIT related arbitration.13 In addition, the model BIT also insists on the aggrieved companies first exhausting host country judicial remedies before resorting to arbitration.14 Given that with regard to most tax disputes in India, years and sometimes decades can go by before the appeals process is finally exhausted, the insistence on exhausting local remedies is going to be a deeply unpopular negotiating point. Another flashpoint in the imposition of status over contracts in the domain of capital gains relates to the area of amalgamations and conversions. By amalgamations, I refer to the merger of one company (the amalgamating company) into another (the amalgamated company) such that the amalgamating company goes out of existence, the assets and liabilities of the amalgamating company are transferred to the amalgamated company and the shareholders of the amalgamating company acquire shares in the amalgamated company. In India, amalgamations take place under the Companies Act 2013 by way of a court order.15 In the case of amalgamations, there are two points at which a tax on capital gains is potentially applicable. The assets of the amalgamating company become the property of the amalgamated company. Further, the shareholders of the amalgamating company receive shares in the amalgamated company. The issue is whether such a movement of assets would result in a transfer for purposes of the ITA. Though there is some conceptual confusion on whether the two events described above can result in a transfer for income tax purposes, the Supreme Court had no such 12 For

more information on the Cairns BIT Arbitration, see http://investmentpolicyhub.unctad.org/ ISDS/Details/691. 13 Article 2.6(iv) of the Model BIT. 14 Article 14.3 of the Model BIT states: ‘The Investor or Investment must first submit its claim before the relevant domestic courts or administrative bodies of the Host State for the purpose of pursuing domestic remedies.’. 15 The National Companies Law Tribunal (NCLT) has the mandate to sanction scheme of amalgamation on the receipt of an application under section 230–232 of the Companies Act, 2013.

70

6 Capital Gains

qualms and readily recognised a transfer when the shareholders of the amalgamating company received shares in the amalgamated company on the extinguishment of their shares in the amalgamating company.16 Transfer has been defined broadly in the ITA to include extinguishment of assets. The amalgamation provisions in the ITA allow transfer of assets between companies as well as the exchange of one company’s shares for another without payment of tax.17 The purpose behind the amalgamation provisions is to allow companies in a poor financial situation to merge into companies with a healthier bottom line so as to take advantage of scale and better management. However, the tax free status of amalgamations lends itself to abuse by companies intending to use the amalgamation provisions as a cloak to obtain tax benefits. Amalgamations are contractual devices used by companies and shareholders to structure their business operations but revenue authorities are concerned, justifiably, that amalgamations can lead to tax abuse. For example, two companies might agree to an amalgamation where, in addition to the issuance of shares by the amalgamated company to the shareholders of the amalgamating company, the very same shareholders also receive cash from the amalgamated company. When the amalgamating company’s shareholders receive cash from the amalgamated company, the arrangement starts looking like a straight sale of shares by the shareholders rather than an amalgamation. The more the proportion of cash in the deal, the more the suspicion that a sale of shares is being dressed up as an amalgamation. Not surprisingly, the Gujarat High Court has held that any transfer of cash by the amalgamated company to the amalgamating company’s shareholders will be fatal to the tax free status of the amalgamation.18 Another example of abuse relates to the ITA allowing loss making companies to carry forward their losses and set these losses against their taxable income for a number of years. A financially successful company might amalgamate into a loss making company solely in order to use the losses of the amalgamated company to reduce the taxable income of the amalgamated company going forward. The ITA now has detailed provisions that prevent such loss shopping except in genuine cases of amalgamations.19 The amalgamation provisions have resulted in some ingenious arguments by tax lawyers. Once such case is the AAR decision in Banca Sella S.p.A.20 This case involved an amalgamation of an Italian company with another Italian company, with the only Indian connection being that the amalgamating Italian company had assets situated in India. The lawyers for the amalgamating company argued that the amalgamation must receive tax free treatment for the amalgamating company under section 47(vi), even though the amalgamated company is not an Indian company, and therefore the conditions mentioned in section 47(vi) are not satisfied. In support of 16 CIT

v. Grace Collis (Mrs.) and Others, (2001) 3 SCC 430. 47(vi) and 47(vii), ITA. 18 CIT v. Gautum Sarabhai Trust, 1988 SCC OnLine Guj 156. 19 Sections 72A, 79, ITA. 20 2016 SCC OnLine AAR 12. 17 Section

6 Capital Gains

71

their argument, the lawyers pointed to a non discrimination provision in Art. 25(1) of the India - Italy DTAA: Non-discrimination-1. The nationals of a Contracting State shall not be subjected in the other Contracting State to any taxation or any requirement connected therewith, which is other or more burdensome than the taxation and connected requirements to which nationals of that other State in the same circumstances and under the same conditions are or may be subjected.

The lawyers argued that Italian amalgamating companies must be treated on par with Indian amalgamating companies and therefore the amalgamation must result in tax free consequences for the Italian amalgamating companies. The AAR agreed with the lawyers.21 It is unclear how the AAR came to this conclusion. The non discrimination provisions apply to ensure parity in tax matters with residents only under the same circumstances. Here the amalgamated company is a foreign company and the section 47(vi) exemption disqualified such amalgamations for Indian resident companies as well, i.e. if an Indian company amalgamated into a foreign company, it would lose the benefit of a tax free amalgamation. In such circumstances, it does not matter if the amalgamating company is an Indian or a foreign company; the amalgamating company will not get the benefit of the section 47(vi) exception if it amalgamates into a foreign company. There does not appear to be any issue of non discrimination here. Nevertheless the AAR was equally impressed with another legal impediment to the taxation of capital gains in this case. Since the amalgamating company did not receive any consideration, capital gains, by definition, cannot be calculated for the transfer of assets from the amalgamating company to the amalgamated company, and therefore the capital gains charge itself fails. Therefore the AAR declared that the amalgamating company cannot be subjected to a capital gains charge in any case because there is no consideration.22 The AAR also stated that transfer pricing provisions will not apply since there is no income tax charge in the first place.23 What’s interesting about this part of the decision is that the AAR decided on the fundamental facets of income tax in a somewhat summary fashion. It is not at all clear that if there is no consideration received, the calculation mechanism fails entirely and therefore, following the Srinivasa Setty24 principle, there is a total collapse of the income tax charge itself. Suppose I transfer property to another person for a zero sum of money. Does this mean there is a total failure of the capital gains calculation mechanism? One can plausibly argue that all this means is that there are no capital gains to tax, which is a result identical to a situation in which the consideration received is equal to the cost of acquisition. It is no one’s argument that in the latter case, the calculation mechanism has failed. The Srinivasa Setty principle should not be extended beyond its scope. It is meant to cover those cases where it is impossible to calculate capital gains. For example, it 21 Ibid,

para 9.

22 Id. 23 Id

para 15. v. B. C. Srinivasa Setty, (1981) 2 SCC 460.

24 CIT

72

6 Capital Gains

is logically impossible to figure out the cost of self generated goodwill. The nature of self generated goodwill does not lend itself to the possibility of costing it for capital gains purposes. Hence the goodwill capital gains mechanism has failed. This idea should not be extended to situations where no consideration flows between the parties but where there is nothing in the facts of the case to prevent the possibility of a capital gain. Since the AAR judged the amalgamation situation to be a total failure of a tax charge, they declined to address transfer pricing issues. However a transfer pricing analysis would have been appropriate here, as the transfer of property was between associated enterprises. The Banca Sella S.p.A. decision provides a measure of flexibility to foreign companies with Indian assets deciding to restructure their business. To the extent allowed in the Banca Sella S.p.A. decision, contracts would continue to hold sway due to the manner in which the capital gains provisions have been interpreted in the ITA and curiously enough, the impact of the non discrimination provisions in many of the DTAAs entered into by India. In addition to amalgamations, another increasingly common manner of restructuring businesses is to convert companies into limited liability partnerships (LLPs). LLPs can take advantage of limited liability for all its members whilst also being taxed as a partnership, with the consequence that the income earned by the LLP is taxed only once. Conversions are therefore a typical use of a voluntary arrangement among shareholders to achieve tax efficiency. The shareholders of the company mutually arrive at a plan to convert the company into a LLP. However, there could potentially be a tax cost to the conversion since there is, on the face of it, a transfer of assets from the company to the LLP. If it is indeed the case that there would be a tax on conversion, then the tax law is imposing a cost on the voluntary restructuring of one’s business. Section 47(a)(xiii)(b) of the ITA recognizes such a possibility and preempts a tax charge by providing tax free treatment if certain conditions are satisfied. Two of these conditions are particularly onerous, as in many mid size companies would struggle to fulfill these conditions: the company’s total sales, gross receipts or turnover should not have exceeded sixty lakh rupees in any of the three years preceding the year of conversion, and the value of the company’s assets in any of the three years preceding the year of conversion should not have exceeded five crores. The repercussions of not fulfilling the ITA conditions for a tax free transfer were examined in Mumbai ITAT decision in ACIT v Celerity Power LLP.25 The taxpayer had converted from a company to an LLP but had failed the condition related to the total sales of the company before conversion. The taxpayer’s turnover was more than sixty lakhs in the year before the conversion. The assessing officer imposed a capital gains tax based on the difference between the market value and the book value of the company’s assets at the time of the conversion. The taxpayer argued that in a conversion, there was no transfer in the first place, and therefore there was no consequent question of imposing a capital gains tax.

25 MANU/IU/1165/2018.

6 Capital Gains

73

The ITAT rejected the argument that there was no transfer for purposes of the ITA. One of the factors that led to the ITAT recognising a transfer was that section 47(a)(xiii)(b) and the memorandum accompanying its introduction in the Finance Act, 2010 envisage implicitly that the normal tax rules would recognise such a conversion as a transfer subject to the enumerated exception in 47a(xiii)(b).26 While this would be the revenue’s interpretation, it is by no means the only interpretation possible. It is by no means a fait accompli that any transaction that has been sanctified by section 47 as a tax free transaction under certain conditions is ipso facto a transaction if those particular conditions are not fulfilled: these transactions by their very nature might not be taxable under the ITA. Nevertheless the ITAT found that there were no capital gains in the facts of the case because the assets of the company had been transferred at book value during the conversion.27 Therefore consideration (the monetary value received by the company on the transfer of assets) equaled cost (the monetary value spent by the company on acquiring the assets); consequently there was no capital gain. The revenue’s attempt to ascribe a fair market value to the assets was rejected by the ITAT because in the capital gains context, the full value of consideration received by the taxpayer need not always reflect the fair market value of the asset.28 Could the revenue impose a fair market value on the basis of transfer pricing provisions? As seen below, a transfer pricing analysis makes use of fair market values in related party transactions but it is difficult to see how transfer pricing rules can apply to conversions.

26 Ibid,

paras 11 and 12. para 17. 28 Id para 16. 27 Ibid,

Chapter 7

Transfer Pricing

Companies and persons related to each other are able to contract with each other to manipulate the prices of goods and services exchanged between them. These manipulated prices would ordinarily be of no concern to the tax revenue except when such manipulation results in less taxes owed compared to situations where goods and services are valued among unconnected parties. The simplest example of this manipulation and the consequent tax consequences would be if one person sold property to a related person at less than fair market value. It is obvious that the undervalued property sale would be subject to a capital gains tax that would be lower than what would have been imposed if the transaction had occurred at fair market value, as would have been the case if the parties were parties unconnected to each other or to use another parlance, dealing with each other at arm’s length. It is important to remember here that a transaction can be reported to tax authorities at less than fair market value among unconnected parties as well. Such parties might decide mutually to report a sale and purchase of property say a piece of real estate, at less than market value and at the same time exchange more money than what is reported. This would simply be a case of black money generation i.e. money that is not accounted for within the tax system. While black money is a problem for the integrity of the tax system, transfer pricing is not per se concerned with the generation of black money. It might well be the case that two unconnected parties report the transfer of goods or services between themselves at below market prices and the money transfer between the parties conforms exactly to their reported positions. The most common reason for related parties transferring a good or a service at below market prices in a transfer pricing situation relates to international tax arbitrage. A multinational with group companies in various jurisdictions can resort to transfer pricing, as explained below, to reduce its global taxes without resorting to hoarding black money. A simple model of transfer pricing involves a parent company of a multinational in a high tax jurisdiction with a subsidiary company in a low (or no) tax jurisdiction. If the parent company were to sell goods and services directly to customers, it would be subject to a substantial tax in its home jurisdiction. Instead the parent company can sell the goods to its subsidiary at a low price and the subsidiary © The Author(s), under exclusive licence to Springer Nature India Private Limited 2020 N. Nuggehalli, International Taxation, SpringerBriefs in Law, https://doi.org/10.1007/978-81-322-3670-2_7

75

76

7 Transfer Pricing

resells the goods at the price the goods can fetch in the international market. The parent company will suffer a much lower tax impact because of the low sales price. The subsidiary will be subject to low or no taxation on the resale if it is in a zero tax haven or a country with low tax rates. The overall result of the ‘transfer pricing’ between the parent and the subsidiary is that the home country of the parent company is denied the tax revenue it would have received if there were no transfer pricing. Given the pervasive nature of transfer pricing, it is a surprise that until 2001, India did not enact a systematic code to deal with transfer pricing. In the absence of transfer pricing law, contractual arrangements would dominate, and dominate in the most invidious way possible, where the most important marker of a commercial relationship, the price, is subject to contractual manipulation. The legal regime relating to transfer pricing in India seeks to remedy the domination of contact through a statutory structure that allows related parties (or associated enterprises as the ITA terms it) to contractually determine the price between them in an international transaction as long as it is an arm’s length price and is determined through one of the methods prescribed by the regulations issued under the statutory provisions of the ITA.1 The five methods are the comparable uncontrolled price method (CUP), the cost plus method, the transactional net margin price method, the resale price method, and the profit split method.2 I won’t spend much time on analysing these different methods as the methods have a detailed and exhaustive history behind them, having been studied by the OECD and adopted by most of the major trading nations around the world. The same can be said about the definition of associated enterprises which lays down rules which are typical for related companies. However, in this area, the government has introduced several deeming provisions in addition to the traditional definition. The ITA begins with a definition of associated enterprises that specifies two enterprises to be related to each other if one participates in the management, control or capital of the other.3 Similarly the ITA also includes another definition of associated enterprises whereby two enterprises are related to each other if one or more persons participate in the management, control or capital of both enterprises.4 The ITA does not define what is meant by participation in the management, control or capital of an enterprise. Instead, it has provided a series of deeming provisions

1 Section 92, ITA states that any income arising from an international transaction shall be computed

having regard to the arms length price. Section 92B has defined an international transaction as a transaction between two or more associated enterprises, either or both of whom are non-residents. Section 92(2) read with 92BA extend some of the ITA’s transfer pricing provisions to ‘specified domestic transactions’. This chapter will not consider the extension of the ITA’s transfer pricing provisions to purely domestic transactions. 2 Section 92C. 3 Section 92A(1)(a). Such a participation can be direct or indirect or through one or more intermediaries. 4 Section 92A(1)(b). Much like the first part of the traditional definition, such a participation can be direct or indirect or through one or more intermediaries.

7 Transfer Pricing

77

which if fulfilled would make it the case that one enterprise is participating in the management, control or capital of another enterprise.5 The deeming provisions on participation in the management and capital of another enterprise are fairly straightforward and consist of some quantitative markers regarding the extent to which one enterprise in invested in the equity or debt of another enterprise. For example, if one enterprise owns shares carrying twenty six percent or more of the voting power in another enterprise, the two enterprises would be deemed to be associated enterprises.6 Similarly, two enterprises would be associated if one enterprise has advanced a loan to the other enterprise that is at least fifty one percent of the book value of the assets of the other enterprise.7 It is when the deeming provisions leave the certainties of quantitative standards and move towards qualitative markers that interpretational problems arise. An example is the deeming provision, section 92A(2)(i) which provides that two enterprises would be considered as associated enterprises if the goods manufactured by one enterprise are sold to the other enterprise or sold to persons specified by the other enterprise and the prices and other conditions relating thereto are influenced by such other enterprise. This provision was the subject matter of a recent judgment by the Chennai Income Tax Appellate Tribunal. In Orchid Pharma Limited v Deputy Commissioner of Income Tax,8 the court had to consider a contractual relationship between an Indian pharmaceutical company and its foreign distributors under which the sale price of the goods sold by the foreign distributors was determined ultimately by the foreign distributors and not the Indian company. The sales through the foreign distributors amounted to less than six percent of the overall sales of the Indian company. The tribunal was faced with the following question. Given the language of the provisions mentioned above, was it not the case that by virtue of the foreign enterprise determining the prices at which the goods are being sold to end customers, the Indian company and its foreign distributors were associated enterprises? If the Indian company and its foreign distributors were considered as associated enterprises, then the prices agreed between the two enterprises would be subject to a transfer pricing adjustment. On the face of it, section 92A(2)(i) requires only that the foreign enterprise influence the price at which the goods are sold to the end customer. This was certainly the case. The upshot of the provision, if interpreted literally, is that because of a contractual arrangement between two parties that made one party determine the final selling price of the products made by another party, the two parties become subject to transfer pricing regulations. Such consequences would be a good example of status over contract. The Chennai tribunal declined to undertake such a literal interpretation. According to the tribunal, section 92A(2) is controlled in its operation

5 The

deeming provisions are found in section 92A(2). 92A(2)(a). 7 Section 92A(2)(c). 8 2016 SCC OnLine ITAT 12571. 6 Section

78

7 Transfer Pricing

by section 92A(1).9 Section 92A(1) broadly focussed on one enterprises controlling another or the same person or persons controlling both the enterprises. Such a control can be manifested legally (de jure) through a dominant stake in the voting power or equity capital or debt capital of an enterprise or factually (de facto) through some other mechanism. Section 92(A)(2) lays out different patterns of these legal and factual dominance. Hence, even if clause (i) of section 92(A)(2) talks merely of influence, what it means is a dominant influence of one enterprise over the prices charged by another enterprises.10 Such was not the case in the facts considered by the tribunal where, as mentioned above, the foreign enterprises influenced the prices not more than six percent of the total sales of the Indian company. The concern at the heart of the transfer pricing debate in India and elsewhere is the idea of an arm’s length price. As long as the transfer pricing rules focus on an arm’s length price, they inevitably cede some ground to multinationals in fixing the prices between their associated enterprises. Perhaps this is unavoidable in international commerce but one must be clear that the very notion of an arm’s length price is in favour of contractual independence in the contract v status debate. Why so? The arms length price aims at the price that two parties would arrive at if they were not related to each other. The focus is on what the two parties would do and not some objective price independent of the intentions of the two parties. The transfer pricing law is a perfect example of the tussle between contract and status in tax law, and one can observe the twists and turns that are characteristic of this struggle. In a status free world, associated parties within a multinational group would determine, via contracts, the prices for the various transactions that occur between the parties including sales of goods and provisions of services as well as financial dealings such as loans and guarantees. For the reasons discussed above, complete contractual freedom granted to multinational groups in this area would result in significant tax evasion. As complete contractual freedom is not recommended between associated enterprises, Indian tax law provides for a constrained contractual freedom, by providing for how prices charged between associated enterprises can conform to an arm’s length price. The guidelines regulating the arms length price is the first instance of an intrusion of status into the field of contracts. The guidelines allow for some extremely limited contractual variation from the prices determined under the guidelines. To finesse the impact of the guidelines, and to pre-empt disputes about the accuracy of the arms length prices, Indian transfer pricing regulations also provide for safe harbours-arms length prices that will not be challenged by the revenue. Safe harbour prices are a sophisticated form of status imposition on contracts. So far, safe harbours have not been popular with the taxpayers, and one of the reasons is that the prices deemed safe from revenue challenge are too far in range from the prices the taxpayers would consider as arms length prices.11

9 Ibid,

para 10. para 17. 11 Vikas (2019). 10 Ibid,

7 Transfer Pricing

79

In a twist to the traditional status v contract debate, the transfer pricing regulations, in a nod to global trends, have provided for advance pricing agreements (APAs) between the revenue and the multinational companies under which the contracting parties agree on the arms length prices or on a formula to determine the arms length prices. The APAs have found more success than the safe harbour rules, although the number of APAs signed is not significant. The presence of APAs shows the limited role of rule based valuations in the transfer pricing area. In other words the imposition of status can only go so far in the area of transfer pricing and even the government has to resort to contracts in order to ensure certainty in transfer pricing matters. So far, the discussion has proceeded on the basis that the transfer pricing guidelines attempt to impose a certain pricing status through their guidelines on various arms length pricing methods although the APAs are recognition of the fact that imposition of external standards in this area is of limited significance. However, there are other ways though which the revenue has attempted to impose status over contracts in transfer pricing matters, and some of these attempts have received judicial benediction. Two of these attempts will be discussed in this chapter. The first attempt relates to what is termed by revenue and taxpayers alike as advertising, marketing and promotion expenses (AMP expenses). These expenses arise when one enterprise licenses its trademark or some other intellectual property to an associated enterprise. When the licensee enterprise, in the process of exploiting the intellectual property, promotes the intellectual property through various means, a question arises as to whether the licensee enterprise ought to be rewarded for its efforts in promoting the intellectual property of the licensor enterprise, even though such a reward is not provided for contractually between the two associated enterprises. The first case on this point is the well known LG Electronics case.12 LG, a Taiwanese electronics giant, incorporated a fully owned subsidiary in India (LG India) in order to exploit the Indian consumer market. LG India sold LG branded electronic items to consumers in India. LG India spent a considerable amount of money promoting LG branded products to potential Indian customers. The revenue believed that expenses incurred on such promotional activities were in excess of what would be normally incurred by domestic firms and characterised a portion of these expenses as expenses on disguised brand building services undertaken by the subsidiary for its parent company. Since these brand building services had not been compensated by the parent, the revenue sought to benchmark the hypothetical price by taking into account the promotional expenses incurred by comparable domestic firms, and then deciding on an appropriate mark up on the benchmarked price. The revenue’s efforts to introduce transfer pricing in the LG case is a remarkable exercise of extending the reach of the transfer pricing statutory guidelines. Notice that the LG case is not a case of the revenue challenging contractual means of avoiding taxation. On the contrary, in the LG case, there was no express contract between the parent and the subsidiary for the provision of AMP services. The revenue argument argued that the transfer pricing guidelines would apply even in the absence of an 12 M/s LG Electronics India Pvt. Ltd. v Assistant Commissioner of Income Tax, MANU/ID/0036/2013.

80

7 Transfer Pricing

express contract for the provision of brand building services between two associated enterprises. In order to succeed in its argument, the revenue had to argue that the ITA rules on transfer pricing applied to the LG situation. Section 92 of the ITA states categorically that the transfer pricing regulations apply only to international transactions. The legislation intent is clear: the transfer pricing rules are to apply not to all kinds of international states of affairs or activities but only to a subset of these, namely international transactions. The definition of an international transaction, contained in section 92F(v), is inclusive in nature and provides that a transaction includes an arrangement, understanding or action in concert, whether or not such transactions are formalised or in writing, and whether or not such transactions are intended to be enforceable by legal proceedings. In LG Electronics, the majority judgement held that the promotional efforts of the Indian subsidiary on behalf of its parent demonstrated the existence of a tacit agreement between the two associated enterprises.13 There was an ‘open blue sea’ policy in place set at the group level according to which the companies in the LG group were going to market themselves into a dominant position and LG India was integral to the effort.14 The LG group had a global marketing strategy and its worldwide subsidiaries were part of an integrated plan of marketing. For example, LG Singapore was in charge of marketing for a region that included India.15 The ITAT therefore drew a straight line of inference from the marketing efforts of the Indian subsidiary to the conclusion that there existed a tacit agreement between the Taiwanese parent and the Indian subsidiary to contribute to the global LG brand. It appears from the majority judgement that the tacit agreement is an objective inference from the facts of the case rather than an inquiry into the subjective intentions of the parties involved. One must note that the objective inference here is with respect to the very existence of an agreement and is not merely an inference relating to the terms of an agreement, although some lawyers might claim there is a thin dividing line between the two. Nevertheless, an objective inquiry such as the one endorsed by the majority puts considerable pressure on multinational groups for the actions of their constituents might come under the transfer pricing scanner unintentionally in the guise of a tacit agreement. The minority judgement refused to recognise an agreement in the facts before it.16 Although it did not present its analysis in terms of an objective and subjective inquiry, the minority judgment has tried to carve out a subjective space for contracting parties in the transfer pricing arena. If two parties did not intend consciously to associate themselves in a mutually acceptable manner, the law cannot impose a tacit association on the parties based on the facts of their interactions. This latter albeit minority interpretation of the meaning of an international transaction in the LG case was in fact accepted by the Delhi High Court in the Maruti 13 Ibid, 14 Ibid,

para 8.9. para 11.1.

15 Id. 16 LG

Electronics, para 32 onwards (page 66–103), minority decision by Mr. Hari Om Maratha.

Reference

81

Suzuki case.17 In Maruti Suzuki, the Delhi High Court was faced with a similar set of facts: an Indian subsidiary incurring expenditure on marketing products that invoked the brand of its foreign parent. The Delhi Court came down firmly on the side of the minority judgment in the LG case. The Court mentioned that a quantitative test of transfer pricing i.e. a test that draws transfer pricing inferences purely from an inquiry into the supposedly excessive nature of a subsidiary’s marketing expenditure has no statutory benediction in Indian tax law.18 The revenue was attempting to concoct an international transfer pricing transaction out of a quantitative inquiry but this was putting the cart before the horse. The transfer pricing inquiry must first identify an international transaction following which an inquiry into the reasonableness of the price or expenditure associated with the international transaction can be made. What a transfer pricing inquiry ought not to do is decide on the reasonableness of an item of price or expenditure and infer an international transfer pricing test as a consequence. The area of transfer pricing has some additional quirks that have some relevance to the contract v status debate. So far the debate has been about the ability of transfer pricing legislation to counteract the prices set contractually among related parties. However, the law relating to transfer pricing can intervene in the contract v status debates through two other mechanisms. One relates to setting out safe harbour prices within which related parties can determine the prices at which they contract with each other. Indian regulations have provided for safe harbour prices but these prices have been perceived as not bringing much succour to related parties contracting with each other. The price range is much too narrow for the parties to exercise much autonomy in deciding on the prices applicable to their dealings. The other legal mechanism that intervenes in the contract v status debate is a contractual mechanism normally negotiated between a multinational group and the state: advance pricing agreements or APAs. An APA allows for prices and the mechanism to determine prices between related parties to be determined in advance. India has entered into a number of APAs with multinational groups in the recent past. APAs appear to be more popular compared to the safe harbour rules. APAs can also be negotiated between countries; such APAs determine an agreed upon mechanism and standards to determine prices between related parties. Recently, India and the United States have signed an APA. An APA is a curious anomaly in the contract v status debate, for it confounds the distinctions between contract and status. An APA, whether of the unilateral or the bilateral form, is a contract that trumps over status (transfer pricing statutes and regulations). However, the contract is not between private parties; by its very nature an APA involves public representatives as one of the contracting parties. An APA is an example of the government seeking to use the contractual mechanism to regulate the application of its own rules. Governments are willing to go down this path because it reduces their tax administration costs.

17 2015 18 Ibid,

SCC OnLine Del 13940. paras 71–73.

82

7 Transfer Pricing

Reference Vasal V (2019) Transfer pricing: advance pricing pact vs safe harbour provision, livemint, 10th May, 2019. Available at https://www.livemint.com/money/personal-finance/transfer-pricing-advancepricing-pact-vs-safe-harbour-provision-1557417827008.html

Chapter 8

General Anti-avoidance Rules (GAAR)

Introduction The GAAR comprise the most recent developments in the status v contract battle described in the preceding chapters. Two strategies are used by governments the world over to impose status over contracts. First, governments introduce deeming provisions in tax statutes to ensure that status dominates contracts through a legal fiction. In some instances, such as the provisions relating to capital gains in the ITA, governments make such deeming provisions retrospective in nature, thus impacting contractual arrangements made by parties who were unaware of, and therefore unprepared for, the deeming provisions that changed their legal position. Second, if the first strategy fails, governments recharacterise the contractual relationship using a broad spectrum set of rules called ‘General Anti- Avoidance Rules’ or GAAR. Over the last decade, the Indian government has used the first strategy extensively, as discussed in Chaps. 4, 5, 6 and 7. The tax provisions relating to FTS & Royalty as well as the area of capital gains and transfer pricing demonstrate the Indian government’s attempt to establish, through legislation, the dominance of status over contract. Recently, the Indian government embarked on the second strategy. The Indian GAAR is the last throw of the dice in the status v contract battle. Even before the BEPS discussions gained a foothold in international tax law reform, India was seriously considering a more rigorous legislative approach to tax avoidance. While the immediate trigger for India’s reform process was the Vodafone case, the legislative process itself was a culmination of the revenue’s disenchantment with the Supreme Court’s decision in the Azadi Bachao case, which limited the somewhat open ended language of Chinnappa Reddy in the McDowell case. To understand GAAR in its context, one must revisit the judicial attitude towards tax avoidance, for if the judicial response to tax avoidance were to be considered as adequate, there would be no reason for the government to initiate a separate legislative response to tax avoidance. In Azadi Bachao and in Vodafone, the courts

© The Author(s), under exclusive licence to Springer Nature India Private Limited 2020 N. Nuggehalli, International Taxation, SpringerBriefs in Law, https://doi.org/10.1007/978-81-322-3670-2_8

83

84

8 General Anti-avoidance Rules (GAAR)

had to adjudicate on the extent to which legislative tax provisions could be interpreted to extend the government’s authority to tax. In Azadi Bachao, the Supreme Court was asked if the government had the authority (even if it was reluctant to exercise it) to reject or look beyond the legal residency of a Mauritius company even if the Mauritius company had fulfilled all the requirements of obtaining Mauritian residency as per the India-Mauritius tax treaty. The Supreme Court replied in the negative, even though typically the Mauritius company was in Mauritius mainly for the reason that its shareholders intended to use Mauritius as a tax saving jurisdiction. The Supreme Court’s point was that the Mauritius route was in fact envisaged by both parties as a jurisdiction through which global investment would flow into Indian markets. Consequently the tax saving intention of the incorporating parties was something that was either not a problematic factor or was something that the tax treaty already took into account. In Vodafone, tax savings was certainly one of the primary reasons for the parties choosing a Cayman Islands company to effectuate the indirect transfer of Indian assets but in the face of the clear provisions in section 9 of the ITA, the court was inclined to accept the legitimacy of the transfer. Both Azadi Bachao and Vodafone demonstrate one major issue with the judicial interpretation of tax statutes: unless the tax legislation expressly asks the courts to take into account an intention on the part of a taxpayer to avoid tax, the courts in India and in the United Kingdom are reluctant to do so, except in certain limited circumstances. The reasons for the courts reluctance to take into account taxpayer intentions is not far to seek. The problem lies in the manner in which the courts have approached statutory interpretation of tax statutes in both the United Kingdom and following the UK lead, in India. While a literal interpretation of tax statutes would be indifferent to taxpayer intentions, a purposive interpretation of tax statutes, such as the one preferred by Lord Hoffman, would also not be capacious enough to include taxpayer intentions to avoid taxation.1 To consider the implications of this point, one must return to the Barclays case.2 In Barclays, the taxpayers made a payment for depreciable machinery, thus entitling the taxpayers to depreciation but they also structured their transactions in such a way that the payee placed the amount received within the control of the taxpayer’s associates. The House of Lords held that the taxpayer had fulfilled the requirements of the tax statute by making a payment, at which point the taxpayer was entitled to depreciation, without regard to what transpired after the payment. The purpose of the depreciation provisions was satisfied as long as there had been an actual investment in depreciable property by the taxpayer. The taxpayer had indeed entered into a series of transactions with an intention to reduce its taxable income but that intention was not relevant to either a literal or a purposive interpretation of the tax provisions in play. Unless the tax provisions explicitly state so, the judiciary in the United Kingdom does not tax taxpayer intentions.

1 Please 2 See

see above discussion of Lord Hoffman’s views on tax avoidance in Chap. 3. above Chap. 3 for discussion on the Barclays case.

Introduction

85

One way to make taxpayer intentions relevant in an interpretation of a tax statute is to use the form v substance argument. The revenue might argue that the transactions of the taxpayers, despite the characterisation provided to the transactions by the taxpayer, can in substance be characterised differently for tax purposes. Recall that this was the argument used by the revenue in the Duke of Westminster case, where the revenue argued that the so called annuities paid by the taxpayer to his gardener are actually wages for tax purposes. For the form v substance argument to work, it has to first make a case for rejecting the taxpayer characterisation of a transaction or a series of transactions. The revenue has to argue that there was something improper or dodgy about the taxpayer characterisation, which is the reason why it has to be superseded by an alternate characterisation.3 Arguing that the taxpayer characterisation is improper is effectively arguing about the taxpayer intentions to avoid tax. Therefore, seeking to ignore the taxpayer characterisation is another way of making taxpayer intentions relevant to interpretations of tax statutes. Despite its air of objectivity, the form v substance test works by invoking taxpayer intentions to avoid taxation. The English and the Indian judiciary have consistently rejected the form v substance approach. The Vodafone case is an example of an instance where in substance the control over Indian assets was being transferred from one party to another but the form of the transfer brought it outside the Indian tax jurisdiction and hence not subject to Indian tax. The taxpayer’s choice of form was intentional and in pursuance of a strategy to avoid tax and could be disregarded only if such intentions could be taken into consideration while applying the tax provisions. At the time Vodafone was decided, taxpayer intentions to avoid tax were not a factor playing any statutory role in the imposition of taxes. The Indian GAAR has made the intention to avoid tax relevant in the imposition of tax.

The GAAR Provisions In 2012, following an international trend in countries enacting or in the process of enacting broad anti-avoidance legislation (examples: United Kingdom and South Africa), the Indian government produced its first draft of a GAAR. The draft went through several versions and was postponed twice, with the consequence that GAAR is operational only from April 1, 2017. The GAAR is the pinnacle of India’s efforts to impose status over contract. The key to GAAR’s application is the idea of an ‘impermissible avoidance arrangement’.4 As it was originally defined, it designated any arrangement that had one of its main purposes as the attainment of a tax benefit as an impermissible transaction. There was much opposition to this definition as it was felt that it would have caught many otherwise legitimate transactions that would have also had tax benefits. After 3 Note

that one is not arguing here that the taxpayer has committed a fraud or that the transactions have no business effect whatsoever. 4 Section 96, ITA.

86

8 General Anti-avoidance Rules (GAAR)

a hue and cry from the business community, this definition was modified by the government in the second iteration of the GAAR. The new iteration of the GAAR states that an impermissible avoidance arrangement refers to an arrangement, the main purpose of which is to obtain a tax benefit. On the face of it, this is a welcome refinement of the most critical definition in GAAR. However, as with most other Indian legislative measures, ambiguity lurks around the corner, for the definition also provides that an arrangement shall be presumed to have its main purposes as obtaining a tax benefit if the main purposes of a step in or a part of the arrangement is to obtain a tax benefit.5 This wrinkle to the definition of an impermissible arrangement in its effect is not very different from a definition that catches an arrangement which has one of its purposes as tax avoidance. Will it not be open to the revenue to isolate one part of a perfectly legitimate transaction and argue that it was for the purposes of obtaining a tax benefit? If the revenue does so, the overall purpose of a business arrangement is irrelevant as long as the revenue is able to plug away at some portion of it. This is a recipe for tax litigation. Once it is determined that an arrangement has its main purpose of obtaining a tax benefit, section 96 provides that the arrangement’s branding as impermissible for GAAR purposes will occur if the arrangement has, in addition, one of four characteristics that revenue officials in India (and across the world) usually consider as suspicious. In other words, section 96 defines an impermissible avoidance arrangement as an arrangement that has its main purpose of obtaining a tax benefit and in addition, contains one of four suspected characteristics. The first suspect characteristic is related to the rights and obligations created between two persons. If these rights and obligations are of the sort that won’t normally be created between persons dealing with each other at arm’s length, such rights and obligations would be suspect from GAAR’s point of view.6 There are several aspects of the characteristic that are worth mentioning. The very existence of this provision in the GAAR chapter might be considered as surprising because there are very detailed provisions already in place regarding transfer pricing in chapter X of the ITA. However there is one crucial difference between Chapter X and these GAAR provisions. Chapter X is concerned with rights and obligations created between parties related to each other in some way. These GAAR provisions on the other hand are meant to address situations where two unrelated parties are creating rights and obligations that two unrelated parties would normally not enter into. Recall the Barclays case. Even though the lessor and the lessee were unrelated parties, it is arguable that the transactions they entered into with the circular transfer of money were something that unrelated parties would normally not enter into. I write that it is arguable because one is not sure how to establish conclusively that two unrelated parties normally would not enter into a certain relationship. One has to perhaps allude to normal commercial practice to establish the alleged abnormality of the relationship between the parties.

5 Section 6 Section

96(2), ITA. 96(1)(a).

The GAAR Provisions

87

The second trigger is that the arrangement results, directly or indirectly, in the misuse, or abuse, of the provisions of the ITA.7 On the face of it, this triggering factor is one of the most ambiguous provisions in the tax legislation. Yet, India is not alone in introducing it. Both South Africa and the United Kingdom use the language of abuse/misuse of legislation in their respective GAAR legislation. One area in which the use of this provision might become controversial is in those cases where the taxpayer takes advantage of provisions that were introduced to incentivise investment. For example, one might argue that generous depreciation provisions are meant to incentivise investment into factories and machineries. If so, if a taxpayer does invest in machinery but uses a Barclays like financing structure, can it be alleged that the taxpayer is abusing or misusing tax provisions? The taxpayer is doing exactly what the legislature encouraged him to do, i.e. spend money on equipment, never mind what happened to the money after that. The revenue might argue that incentivising investment means incentivising real investment where the money does not come back to the investor surreptitiously. If the money does come back to the investor in some circuitous manner, it is arguable that this would be a misuse or an abuse of tax legislation. The third factor is that the arrangement lacks commercial substance or is deemed to lack commercial substance under the GAAR rules, in whole or in part.8 This is a version of the economic substance test that has been adopted so vigorously in the United States. It is not an easy test to implement because the taxpayer is bound to argue that the transactions in which he has participated have some commercial substance to them. Then the question becomes one of degree. To what extent ought a transaction to have economic substance such that it is not considered impermissible tax avoidance? To put it negatively, what deficiencies in the commercial ordering of transactions make it impermissible tax avoidance. The GAAR rules include a set of deeming provisions in section 97 of the ITA that set out the characteristics of the arrangement that would trigger the commercial substance test. Section 97 begins with a general test according to which an arrangement does not have commercial substance if the substance or effect of the arrangement as a whole, is inconsistent with, or differs significantly from, the form of its individual steps or a part. The general test of commercial substance is a reworking of the form v substance argument discussed earlier in this chapter and in the chapter on the judicial attitudes towards tax avoidance. Recall that the form v substance argument was not meant by the revenue to claim that there was no economic substance at all to a transaction or a series of transactions; rather it was meant to claim that the economic reality of the transaction or a series of transactions for tax purposes was different from what was being claimed by the taxpayer. Section 97 adds two more tests to the general tests described above. The second test identifies certain features of the arrangements that would deem the arrangements as having no economic substance. If the taxpayer puts money into an arrangement that

7 Section 8 Section

96(1)(b). 96(1)(c).

88

8 General Anti-avoidance Rules (GAAR)

comes back to him in some way,9 or if he uses an entity or a person in his arrangements only to obtain tax benefits or to disguise the ordinary tax effect of his arrangement,10 section 97 would deny the commercial substance of such arrangements. The second test’s inclusion of round tripping is intriguing because it changes the complexion of the commercial substance test. When it comes to tax avoidance through arrangements that lack commercial substance, there are two ways in which a person can reduce his taxes. He can pursue a complex scheme that has steps that do not make sense other than for purposes of reduction of taxes. In this scenario, he has money that he has acquired legitimately and has accounted for legitimately under the accounting rules applicable to him or entities controlled by him. Alternatively he can also do something more sinister. He might have access to money acquired illegitimately (proceeds from corruption, drugs, terrorism etc., or just plain tax evasion) that he would like to cleanse i.e. money launder. He can initiate a money laundering scheme under which he undertakes a set of complex transactions that have no purpose other than to camouflage the source and nature of the money and convert, to use a colloquial term, black money into white money. A typical way to do this is to create a money flow that begins with the money launderer and uses a number of transposed entities, called accommodation parties, to hide the ultimate source and the ultimate destination of the money. As an illustrative case, let’s take an example of a company that needs to launder its black money. The company can transfer the money, through surreptitious means, to a number of persons or entities (accommodation parties) and these accommodation parties will in turn invest in the company’s shares at inflated figures, i.e. by subscribing to the company’s shares at a hefty premium. The company’s coffers are now filled with money that has been cleansed of their illegitimate origins as now they are shown as share subscription money. Subsequently the company can even manage to get rid of the accommodating parties by doing a buy back of their shares at nominal costs. The ITA, apart from the GAAR, has provisions aimed precisely at round tripping regimes. Accordion to section 68 of the ITA, if a person is found with financial credits (for example, in the form of share subscriptions or loans) on its books of accounts, and there is no satisfactory explanation given by the person for these credits, the amount mentioned in the books of accounts would be deemed to be the income of the person and will be taxed at a punitive rate. The question of share subscriptions under section 68 has been a source of controversy and litigation. The problem is essentially related to the burden of proof on the part of the revenue and the assessee respectively. Suppose the company is able to identify the shareholders who invested in the company. The identification of shareholders means, at the very minimum, that the shareholders are not just fictitious names. However, the shareholders can be real persons or companies and yet be intermediaries capable of routing spurious money. This might be the case even if these shareholders have invested in the company through banking channels. 9 Such a circular flow of money is called round tripping and section 97(b)(i) read with section 97(2)

have detailed deeming and definitional provisions on round tripping. 97(b)(ii).

10 Section

The GAAR Provisions

89

The section 68 test asks if a person’s explanation of a recorded monetary sum is satisfactory. Is the company’s explanation satisfactory, at least to the extent that the burden of proof shifts to the revenue, if it is able to demonstrate the existence of genuine shareholders and the investment of money through normal banking channels. The leading cases on this point, Divine Leasing11 and Stellar Investment,12 interpreted section 68 in such a way that the assessee’s burden of proof was satisfied in the conditions described above. The Indian government, in response, amended section 68 to make it harder for the assessee to wriggle out of section 68. A proviso to section 68 says that in the case of share subscriptions, the company’s explanation of the genuineness of the sum received will not be satisfactory unless the explanation of the person who invested in the shares is also satisfactory. The round tripping scheme, although it lacks commercial substance, is not a pure tax avoidance scheme like the first one, although it could be argued that tax avoidance is always one of the main purposes in any money laundering scheme. The reference to round tripping therefore brings money laundering schemes to the forefront of the GAAR. How the GAAR would reconcile its focus on traditional tax avoidance with its attack on money laundering schemes is something yet to be tested in India. Does the inclusion of round tripping in the GAAR means the burden of proof in favour of the revenue has been relaxed even further than what is the case now? While the first test under section 97 is a general one, the second test identifies certain suspect features of an arrangement that lacks commercial substance. The third test addresses what might be called the classical case of an arrangement that is economically unsound: it just does not have the effect of an ordinary commercial transaction. Section 97(d) gives a more precise formulation when it states that an arrangement would be considered as lacking commercial substance if it does not have a significant effect upon the business risks or net cash flows of any party to the arrangement apart from any tax benefit that would be obtained by entering into such arrangement. The final trigger for the impermissibility characterisation is the fact an arrangement is entered into, or carried out, by means, or in a manner, which are not ordinarily employed for bona fide purposes.13 The final impermissibility factor appears to be very similar to the first impermissibility factor which focuses on arrangements not normally entered into between parties transacting at arm’s length. Future revenue action and case law will perhaps make things clearer. The GAAR rules contain three different kinds of mitigations meant to lighten the harshness of their application.14 First, income arising out of transfers (of shares or assets) before April 1, 2017 is grandfathered i.e. not subject to GAAR application.15 11 CIT

v Divine Leasing and Finance, 299 ITR 268 (2007). v Stellar Investments 192 ITR 287 (2007). 13 Section 96(1)(d). 14 This paragraph and the succeeding paragraph were published on the blog, IndCorpLaw at http:// indiacorplaw.blogspot.in/2016/07/gaar-and-tax-treaties.html. 15 Central Board of Direct Taxes, “Clarifications on implementation of GAAR provisions under Income Tax Act, 1961”, (January 27, 2017) http://itatonline.org/info/wp-content/uploads/2017/01/ CBDT-GAAR-Press-Release-Circular.pdf. 12 CIT

90

8 General Anti-avoidance Rules (GAAR)

Second, certain transactions, whether before or after April 1, 2017, are in any case kept outside the purview of GAAR.16 For example, certain investments by foreign institutional investors are excluded.17 Similarly, transactions that result in tax savings below three crores are outside the scope of GAAR.18 Finally, a very different kind of mitigation, one that was included because of taxpayer demands, has been introduced. This relates to the process by which a revenue official imposes GAAR. He or she cannot do so without making a reference to the Commissioner of Income Tax.19 If the Commissioner agrees with the GAAR application, the tax payer can appeal to a GAAR approving panel, which consists of a retired or serving High Court judge, a senior revenue official and an independent tax scholar.20

GAAR and International Contractual Arrangements21 The issue going forward for India is whether the GAAR will prove effective in resolving the most egregious international tax avoidance schemes. Note that the Indian government brought in GAAR primarily to address international tax avoidance. International tax issues are complicated by the fact that investors are governed by tax treaties and therefore will seek protection under the relevant tax treaty. The ITA makes it clear that the GAAR will have effect notwithstanding any provisions contained in the tax treaties.22 Since the GAAR is an anti-avoidance provision, its overriding effect would be on anti-avoidance provisions, if any, in tax treaties entered into by India. However, this is precisely where there is cause for confusion. Consider for instance the ‘Limitation of Benefits’ (LOB) clause in tax treaties. Because of tax treaties, investors in India enjoy a favourable tax regime for various items of income such as capital gains, business income, dividends, interest and royalties. As a consequencer, companies began taking advantage of tax treaties without being really resident in the countries with which India had tax treaties. Singapore and Mauritius were jurisdictions that were particularly rife with such shell or postbox companies. To counter treaty shopping by companies, the Indian government negotiated with its treaty partners to introduce ‘Limitation of Benefits’ (LOB) clauses in tax treaties. In August 2005, India and Singapore agreed to include a LOB clause in the DTAA between the two countries. The LOB clause mentions that a company has to invest 16 Ibid. 17 Ibid. 18 Ibid. 19 Ibid. 20 Ibid. 21 This section draws on my work published on the IndCorpLaw blog at http://indiacorplaw.blogspot. in/2016/07/gaar-and-tax-treaties.html. 22 Section 90(2A), ITA.

GAAR and International Contractual Arrangements

91

a certain amount of money in its home country in order to avoid being termed as a shell company, a classification that would result in a loss of tax treaty benefits. For instance, a company resident in Singapore can escape shell company status under the India-Singapore DTAA if it incurs an annual expenditure of at least two hundred thousand Singapore dollars.23 As the name suggests, the LOB clause is a limitation on the benefits that a company can avail under a treaty; however, a rose by any other name would smell as sweet—the LOB is also a safe harbour provision; it signals to companies that as long as they comply with the minimum expenditure requirements of the LOB clause, they need not be bothered about being labelled as treaty opportunists or worse, treaty abusers. Recently India has introduced a virtually identical LOB clause in its treaty with Mauritius, but as a temporary measure.24 India decided to withdraw the generous capital gains exemption it allowed under the IndiaMauritius DTAA (and the India-Singapore DTAA) but has allowed the benefits to continue for a limited period for genuine Mauritius residents i.e. those companies that fulfill the LOB clause. The GAAR has raised doubts about the impenetrability of the safe harbour in the LOB clause. For example, can the revenue continue to claim that a company is making a spurious tax claim as a Singapore resident even though it fulfills the expenditure requirements of the LOB clause in the India-Singapore tax treaty? It would be odd to make such a claim but the vaguely worded GAAR provisions almost invite outrageous claims from the revenue. What makes matters worse is that the ITA is explicit on the following point: if there is a conflict between tax treaty provisions and GAAR, the GAAR would prevail. It’s almost inviting the revenue to play fast and loose with the LOB provisions in tax treaties. The Shome Committee Report on GAAR anticipated this issue and recommended that the government amend the GAAR legislation to clarify that if a company satisfies LOB provisions, it would not be subject to GAAR.25 However, the government has not amended the GAAR rules to this effect. It is incumbent on the Indian government to clarify this issue for foreign companies availing treaty benefits. Until then, the GAAR will hang over international taxpayers like a Damocles sword.

GAAR: A Comparative View General Anti-Avoidance rules are a good example of international convergence of tax avoidance standards. For example, the Indian GAAR language is strikingly similar to the language used in the South African GAAR. Several other countries have also introduced the GAAR; among these are Canada, Australia, the United Kingdom, Hong Kong and Singapore. The Singapore GAAR is particularly instructive as it 23 Article

24A(4)(b), India-Singapore DTAA. 27A, India-Mauritius DTAA. 25 Expert Committee, “Final Report on General Anti-Avoidance Rules (GAAR) in Income-tax Act, 1961” (2012), page 49. 24 Article

92

8 General Anti-avoidance Rules (GAAR)

is an example of a simple GAAR which nevertheless is designed to achieve its objective of capturing abusive transactions. It consists mainly of two paragraphs, one concerned with the conditions for the imposition of the GAAR and the other with conditions for the non-application of the GAAR. The first set of conditions (the conditions necessary for GAAR application) is worded as follows: Where the Comptroller is satisfied that the purpose or effect of any arrangement is directly or indirectly—(a) to alter the incidence of any tax which is payable by or which would otherwise have been payable by any person; (b) to relieve any person from any liability to pay tax or to make a return under this Act; or (c) to reduce or avoid any liability imposed or which would otherwise have been imposed on any person by this Act, the Comptroller may, without prejudice to such validity as it may have in any other respect or for any other purpose, disregard or vary the arrangement and make such adjustments as he considers appropriate.

The second set of conditions (the conditions necessary for the non-application of the GAAR) is worded as follows: GAAR shall not apply to any arrangement carried out for bona fide commercial reasons and had not as one of its main purposes the avoidance or reduction of tax. The Singapore GAAR differs from the Indian GAAR in that the main focus is on the intentions of the taxpayer. The Singapore GAAR accounts for the presence of an intention and for the lack of another intention. The taxpayer must intend to carry out his arrangements for bonafide commercial reasons and must not intend mainly to avoid taxation. Under the Indian GAAR, not only must the revenue establish that the taxpayer intention was primarily to avoid tax but in addition the revenue must also prove that the taxpayer arrangements have one of the features deemed problematic by the GAAR legislation, such as commercially unusual (out of the ordinary) arrangements between two parties. I do not believe that the Singapore GAAR has sacrificed efficiency or coverage in relying mainly on the intention test. It is quite likely that the suspect characteristics identified in the Indian GAAR would in any case be considered by an authority looking into the intentions of a taxpayer. To test this proposition, let us consider the application of the Indian GAAR to the facts of the Barclays case. In Barclays, the taxpayer bought an asset and leased it to another person and through a convoluted series of transactions involving multiple corporate entities one of which was in the Isle of Man, received back the money it paid for the asset in the form of a security deposit by the lessee. If the Indian GAAR were applicable to these facts, the revenue would have to establish that Barclays’ main intention in entering this arrangement (or any part of this arrangement) was to avoid tax and in addition, establish that the arrangement had one or more of the features considered as suspect under the Indian GAAR. In order to infer an intention to avoid tax, the revenue would have to look more closely at the structure put in place between Barclays and the lessee. The revenue might be interested in the following characteristics of arrangement. First, while asset based financing is a perfectly normal commercial arrangement, and it is also reasonable to require a security deposit from the lessee, it is unusual for the leasing contract to be drafted in such a manner that the amount to be received by the seller/lessee is

GAAR: A Comparative View

93

never in reality available to the seller/lessee and is instead routed back on the same day to the lessor’s group. Second, the revenue might argue that because the arrangement resulted in no actual transfer of consideration but only an extraction of tax savings from capital allowances, the taxpayer’s intention was in fact to misuse the capital allowances legislation. In making this argument, the revenue will attempt to ascribe a tax avoidance intention to the taxpayer by ascribing a legislation misusing intention to the taxpayer. Instead of proving the more general (and more difficult) proposition that the taxpayer wanted to avoid tax, the revenue will try to prove that the taxpayer wanted to improperly exploit the shortcoming in a legislation. Third, the revenue might want to prove the tax avoidance intention by assessing the economics of the arrangement. There was no economic substance to the Barclays arrangement as there was no significant effect upon the business risks or net cash flows of any party to the arrangement apart from the capital allowances that would be obtained by entering into such arrangement and shared between the parties. If there is no economic substance to the arrangement, there is a reasonable inference that the taxpayer entered into the arrangement because of the tax savings associated with the arrangement. Finally the revenue might point to the various structural features of the arrangement to advance the argument that the taxpayer intended to avoid tax. In structuring the asset finance arrangement, Barclays interposed two subsidiaries that were not commercially significant to the transaction because the arrangement could have worked without their involvement. One of these subsidiaries was in the Isle of Man, which is a jurisdiction that had no commercial connection to the transaction. The revenue might argue that the addition of commercially superfluous entities is not a bona fide finance leasing arrangement thus demonstrating that the taxpayer’s real intention was to avoid taxation and hide its real intention behind the smokescreen of various corporate entities. As is clear by now, the arguments that the revenue might make, indeed must take, to prove an intention to avoid tax, will often coincide with the characteristics mentioned in the Indian GAAR legislation that the revenue must identify after the revenue has proved the tax avoidance intention. Would it not be much simpler for the GAAR legislation to focus only on taxpayer intentions, much like the approach adopted by the Singapore GAAR? The merits of the Singapore approach can be assessed by considering the landmark case of Comptroller of Income Tax v AQQ and another appeal,26 where the assessee had used a complicated financing mechanism to reduce its taxable income. The financing mechanism used a number of intermediaries to mask what was in effect a circular flow of money that did not affect the economic position of any of the parties involved in the scheme and yet generated interest deductions for the assessee. The Court of Appeals, the highest court of the land, decreed that the financing mechanism was a tax avoidance mechanism under the Singapore GAAR. The financing mechanism was artificial and contrived with one of the intermediaries in Mauritius having 26 [2014]

SGCA 15.

94

8 General Anti-avoidance Rules (GAAR)

no real connection with the taxpayer business needs, the financing arrangement did not generate any real economic risk for the financiers, and the deductions claimed by the assessee did not have any business purpose. Thus the Singapore GAAR enabled the judges to look into the markers of unacceptable tax avoidance that are mentioned in more detailed GAAR such as that found in Indian legislation.

Chapter 9

BEPS, The Principal Purpose Clause and Indian Taxation

The BEPS project is the most significant international trend in the domain of taxation since the advent of bilateral tax treaties. I will discuss the BEPS project in two tranches. The first part will be a brief overview of the BEPS project and its primary features. The second and more detailed part will focus on one of the most interesting and controversial aspect of the BEPS standards: the Principal Purpose (PP) clause and its implications for Indian taxation.

BEPS: Overview and Summary The trend in Indian tax law away from contract towards status was provided an additional impetus by the adoption of BEPS (Base Erosion and Profits Shifting) Action Plans by the OECD. The BEPS Action Plans were developed by OECD after being mandated by the G20 group of countries to develop principles aimed at preventing international tax arbitrage by multinationals. The BEPS Action plans were meant to be principles that would be adopted by each of the countries involved in the BEPS process into their respective legislative and tax treaty frameworks. The BEPS Action Plans had a broad mandate in mind and the final principles developed by the participating countries covered a wide variety of international tax issues ranging from new rules on permanent establishment, hybrid structures, digital services and transfer pricing documentation to binding dispute resolution mechanisms. The principal idea behind the BEPS regime is to ensure, to the extent possible, that the countries in which economic value is created are able to tax the value addition occurring within their jurisdiction. The part of this chapter on equalization levy has appeared in Nigam Nuggehalli, ‘India’s Implementation of the BEPS Project: A Critical Survey’ in Parthasarathi Shome (ed), Insights into Evolving Issues of Taxation (CCH 2016) Part 2.

© The Author(s), under exclusive licence to Springer Nature India Private Limited 2020 N. Nuggehalli, International Taxation, SpringerBriefs in Law, https://doi.org/10.1007/978-81-322-3670-2_9

95

96

9 BEPS, The Principal Purpose Clause and Indian Taxation

The BEPS regime would like to connect value creation and tax jurisdiction in two complementary ways. First, it would like to propose rules that make this connection clearer and stronger. For example, digital sales and services have the tendency to undercut the taxing rights of nations that have a large number of customers paying for these services. The BEPS regime addresses this problem by endorsing principles that explicitly provide such countries with a right to tax digital profits provided certain indicators are met.1 The BEPS regime also provides for a system that makes sure that such indicators are available to revenue authorities in countries. To ensure this outcome, the BEPS regime has imposed more onerous disclosure requirements for multinationals by introducing the Country by Country (CbC) reporting programme.2 The CbC reporting is in addition to the multilateral information exchange agreement promoted by the OECD in order to move to enable an easier and more efficient exchange of information between tax authorities. The BEPS regime does not translate automatically into domestic tax legislation. It is up to the countries to implement the BEPS rules through their domestic tax legislation. In addition, the BEPS initiative also includes, as an integral part, the development of a multilateral tax treaty which would simultaneously amend bilateral treaties to include the new BEPS rules. India has already taken steps to implement the BEPS rules, albeit in a progressive fashion. In the 2016 budget, the Indian government announced plans to impose an equalization levy (discussed below) on non-residents earning online advertising revenues. The Indian government’s move against online advertisement earnings is in line with the issue of the taxation of digital sales that the BEPS regime has addressed in its first Action Plan. The BEPS regime recognised that the sale of goods and services through the Internet needs a radical restructuring of the international tax regime. The current regime, with its emphasis on a strong physical presence of the seller (of services or goods) in the host country, is unable to grant jurisdiction to tax on the host country in the case of online sales conducted remotely through the Internet. To address this problem, BEPS has discussed the idea of taxing non-residents for a ‘significant economic presence’ in the host country which will take into account factors such as the number of customers, volume of sales, amount of data collected etc.3 Ultimately, the BEPS regime decided not to recommend a tax nexus based on significant economic presence because of the complexities involved in the implementation of this standard.4 Despite a tepid reception in the BEPS regime, India amended the ITA in 2018 to include the significant economic presence test as an additional parameter for determining if a foreign enterprise has taxable Indian source income. However, this amendment will have a very limited impact since the overwhelming majority of investment into India comes from countries that have DTAAs with India,

1 See

generally, BEPS Action Plan 1, available at http://www.oecd.org/ctp/beps-actions.htm. generally, BEPS Action Plan 13, available at http://www.oecd.org/ctp/beps-actions.htm. 3 BEPS Action Plan 1, pp. 107–111. 4 Ibid, 137. 2 See

BEPS: Overview and Summary

97

and none of these DTAAs include a significant economic presence as a test for a foreign enterprise’s PE in India. However, the willingness of the BEPS project to consider the significant economic presence test demonstrates that the BEPS project is not averse to a country asserting its right to tax value creation not only because the profit making activity actually occurred within its jurisdiction but also because the customer base for the profit making activity is within its jurisdiction (even if the profit making activity itself is outside its jurisdiction). Therefore value creation as understood in the BEPS regime can tolerate an extension of taxing rights to countries on the basis that any one of the following three connecting factors—the creators of the profit making activity or the profit making economic activity or the consumers of the profit making activity—are within the jurisdiction of the relevant countries. The BEPS regime also discussed the idea of an equalization levy which is a withholding tax on business income earned online, regardless of whether the seller or supplier has a permanent establishment in the host state.5 Once again, BEPS declined to grant an unequivocal recommendation regarding the equalization levy and was content with a statement allowing host states to impose such a levy if they wish.6 Note that the equalization levy is an ad hoc but radical departure from the rule that business income in the host state can be taxed only if there is a physical presence of the person earning the business income. As such, this is the most interesting example of the imposition of status over contract, as regardless of where the parties to a supply of goods and services enter into a contract, and regardless of how the parties structure their contract, the supplier will be taxed in the host state as long as the consumers are resident in the host state. In India, the equalization levy has been implemented in a slightly haphazard manner. First, it has been imposed through a Finance Act (the Finance Act 2016) rather than through an amendment to the ITA.7 Presumably, this route has been taken to avoid a thorny problem of international taxation-any attempt by the host country to expand its jurisdiction to tax non-resident income would potentially contravene its obligations under double tax treaties. For example, in its double tax treaties, India is committed to taxing the business income of non-residents only if the non-residents have a permanent establishment in India. BEPS Action Plan 1 has made it clear that an equalization levy is acceptable only if it does not violate the treaty obligations of countries.8 Further, the levy taxes only a provision of services, not a supply of goods. Only services with two characteristics come within the catchment area of this levy. The service provider must offer advertising opportunities to its customers, although there is a provision allowing the government to extend the equalisation levy to other kinds of

5 Ibid,

115. 137. 7 Chapter VIII, Finance Act, 2016. 8 BEPS Action Plan 1, p. 137. 6 Ibid,

98

9 BEPS, The Principal Purpose Clause and Indian Taxation

services. The advertising opportunities must be online or digitally enabled.9 Finally, the person who pays the service provider must either be an Indian resident carrying on a business or a profession or permanent establishment of a non resident person.10 The equalisation levy as India has envisaged it has content (advertising), delivery (online or digital) and consumer (Indian resident or permanent establishment) limitations. Yet it is a radical departure for Indian tax law. The equalisation levy is plainly contrary to the principle that non-resident business income is taxable in the host state if and only if the non-resident has a permanent establishment in the host state. After the imposition of the equalisation levy, an advertisement service provider such as Google can no longer avoid an Indian tax nexus by entering into contracts with its customers outside India. The situs of the contract is no longer relevant. What is relevant is whether the customer paying Google is an Indian resident carrying on a business or profession in India or is a permanent establishment of a non resident enterprise. The BEPS regime has gained traction in recent times with India signing the Multilateral Tax Instrument (MLI). The BEPS regime envisages that the end result of the BEPS process is a modification of the tax rules applicable worldwide to multinational groups. However, it is at the stage of the modification of the tax rules that the BEPS process runs into an implementation problem. The tax rules that are applicable to multinational companies are mainly contained in bilateral tax treaties. Since there are hundreds of such treaties, each of these treaties have to be amended separately to implement the BEPS action plans. Clearly this process would be cumbersome and time consuming. To address the problem, right from the beginning, the BEPS process rather presciently focussed on its implementation stage through a separate policy document called BEPS Action Plan 15. The BEPS Action Plan 15 envisages a MLI which would be an international legal instrument that would simultaneously modify the extant bilateral tax treaties, thus making more efficient the process of modifying the global tax regime to reflect the changes proposed in the BEPS Action Plans. The MLI is a hybrid international instrument; it attempts to modify multilaterally the different existing bilateral relationships in the tax world. The underlying idea behind this hybrid instrument is that since countries have already developed a tax relationship at the bilateral level with its attendant unique characteristics and peculiarities, it would be prudent to supplement and modify such relationships rather than to supplant them. The MLI works by modifying treaties that it classifies as covered treaties. Covered treaties are those bilateral treaties with respect to which both the contracting parties have agreed to abide by the modifications proposed by the MLI.11 The MLI is flexible in that the signatories to the MLI can pick multiple ways of complying with the MLI 9 Section

163(3) of the Finance Act, 2016 states that the equalisation levy would apply only to specified services. Section 164(i) defines a specified service as ‘online advertisement, any provision for digital advertising space or any other facility or service for the purpose of online advertisement and includes any other service as may be notified by the Central Government in this behalf.’. 10 Section 165(1)(i) of the Finance Act, 2016. 11 MLI FAQ, p. 4, available at https://www.oecd.org/tax/treaties/MLI-frequently-asked-questions. pdf.

BEPS: Overview and Summary

99

as well as opt out of certain portions of the MLI. However, there are certain minimum standards in the MLI from which countries cannot be opted out. One of these mandatory standards relate to the Principle Purpose (PP) clause, which is envisaged by the BEPS Project as an anti-abuse provision. The PP clause, once incorporated into a bilateral treaty, allows the contracting parties to deny treaty benefits if one of the principal purposes of an arrangement entered into by a person claiming benefits under the treaty is to obtain a tax benefit, unless the tax benefit obtained is in furtherance of the object and purpose of the treaty.12 The PP clause is wider than its counterpart in the Indian GAAR. While the Indian GAAR applies only if the main purpose of a taxpayer arrangement is to obtain a tax benefit, the PP clause is applicable as long as one of the principal purposes of a taxpayer arrangement is to obtain a tax benefit. In practice, however, there might not be a significant difference between the two clauses as the Indian GAAR would apply even when a part of or a step in the taxpayer arrangement has the purpose of obtaining a tax benefit.13 The rest of this chapter will discuss the implications of the PP clause in more detail.

The Principal Purpose Clause Introduction One mechanism to shift profits out of a country (the country that is the source of profits, hereafter the ‘source country’) is for an entity to base itself in another country with which it has only a nominal residential nexus (hereafter the ‘residence country’) and take advantage of low or no taxation under the aegis of a tax treaty between the resident county and the source country. These enterprises abuse the treaty process, and one of the most important parts of the BEPS Action Plans is its emphasis on preventing treaty abuse. There are various ways to address the problem of a spurious residency. One way is to insist on the taxpayer fulfilling markers that usually are symptomatic of a genuine residency. For example, the treaty anti abuse provisions can require that the taxpayer must be engaged in an active business in the treaty country or that the taxpayer must be listed on a securities exchange in the treaty country. Some treaties might insist that the taxpayer spend a certain amount of money every year on business related activities in the treaty country. Anti abuse measures that insist on certain concrete markers of a genuine presence in the tax treaty country run the risk that a company might fulfill these markers whilst using the treaty predominantly for tax purposes. Even if a company qualifies as a resident under the domestic tax law or under the treaty, it should not be allowed to enjoy tax benefits accruing under a tax treaty if the only reason it became a resident is 12 MLI

FAQ, p. 9, available at https://www.oecd.org/tax/treaties/MLI-frequently-asked-questions. pdf. 13 See Chap. 8 for an analysis of the Indian GAAR.

100

9 BEPS, The Principal Purpose Clause and Indian Taxation

to exploit the tax benefits arising out of the treaty. To address such situations, BEPS Action Plan 6 has recommended a general anti avoidance rule called the ‘Principal Purposes’ clause (the PP clause), according to which an enterprise would lose the benefits of a tax treaty if one of the principal purposes of the enterprise in obtaining residency in the tax treaty country was to take advantage of a tax benefit arising from the treaty. The BEPS process considers the PP clause as a key element in its efforts to combat tax avoidance; so much so that the signatories to the MLI must, subject to certain limited exceptions, include the PP clause mandatorily in the bilateral treaties modified through the MLI.

BEPS Action Plan 6 and the PP Clause Abuse of tax treaties commonly refers to a process called treaty shopping that occurs when enterprises take advantage of tax treaties in order to reduce or eliminate their tax burden without having any substantial independent reason for being resident in a country with a favourable tax treaty. Before BEPS Action Plan 6, a few countries led by the United States already had anti treaty abuse provisions in place. For example, the India-US DTAA contains provisions aiming to ensure that an enterprise can benefit from the India-US DTAA only if that enterprise has a substantial connection with either India or the United States.14 The idea of a substantial connection is represented in three different ways. First, a majority of the company’s shares ought to be owned by individuals resident in one of the treaty countries.15 Curiously the individuals can be resident either in India or in the United States. For example, a US company with a majority Indian shareholding will be able to protect itself from Indian taxes under the India-US DTAA. In addition to the ownership requirement, the enterprise must not spend a substantial part of its income meeting liabilities that are due to persons who are not residents of either India or the United States.16 This provision is in place in order to ensure that the shareholding requirement mentioned above is not subverted through proxy Indian or US shareholders who are merely conduits for any dividends or interests received by them. Alternatively, the substantial connection can be established by an enterprise conducting an active trade or business in the country in which it is resident.17 Finally, another method of establishing a substantial connection to either India or the United States is for an enterprise’s principal shares to be traded on a recognised stock exchange in either India or the United States.18

14 Article

24. 24(1)(a). 16 Article 24(1)(b). 17 Article 24(2). 18 Article 24(3). 15 Article

The Principal Purpose Clause

101

The LOB clause in the India-US DTAA, in sum, looks at three alternative markers of a substantial connection: the substantial ownership of an enterprise seeking tax benefits must be in the hands of the residents of one of the contracting parties to the DTAA, or the enterprise must conduct an active business in one of the treaty countries, or the enterprise must be listed on a recognised stock exchange in one of the contracting states. Action Plan 6 includes a similar LOB clause, and in this respect, there is no radical departure from the LOB test already employed in tax treaties by India, the United States and other countries.19 In addition to the LOB clause, BEPS Action 6 also recommends the PPT rule, which states the following: Notwithstanding the other provisions of this Convention, a benefit under this Convention shall not be granted in respect of an item of income or capital if it is reasonable to conclude, having regard to all relevant facts and circumstances, that obtaining that benefit was one of the principal purposes of any arrangement or transaction that resulted directly or indirectly in that benefit, unless it is established that granting that benefit in these circumstances would be in accordance with the object and purpose of the relevant provisions of this Convention.20

With the introduction of this provision, the language of general anti avoidance rules (GAAR), flourishing lately in various domestic tax statutes, has now emerged formally in international legal texts, and will become more common in tax treaties worldwide in the future. The importance of such a feature cannot be underestimated because the language used is capacious enough to become a major issue for taxpayers worldwide. The MLI makes it the case that every so called covered treaty will include the PP clause, except in certain limited circumstances. In effect the PP clause is likely to be included in most of the important tax treaties globally.

The PP Clause and the Interpretation of Tax Statutes Indian case law on tax avoidance has struggled to deal with innovative techniques of tax planning that, while avoiding taxes by adhering to the letter of the law, violate the principle or the policy or the spirit behind the relevant tax legislation. As noted earlier, the Indian judicial attitude towards tax avoidance went through several stages. At the first stage, the courts stated that the well known Duke of Westminster doctrine would be applicable according to which a tax can be imposed only if the letter of the law covers the transaction in issue, even if the taxpayer has entered into a transaction in order to escape taxability under the law. At the second stage, the courts changed tracks and said that a colourable transaction, i.e. a transaction conducted with the intention to avoid taxation, would not be given judicial recognition. This stage was superseded by a third stage, in which the courts arrived at an intermediate position, where a taxpayer intention to avoid taxation would not taint transactions if the clear 19 See

discussion in BEPS Action Plan 6 Final Report, Part A (1)(a). Action Plan 6 Final Report Part A(1)(a), p. 55.

20 BEPS

102

9 BEPS, The Principal Purpose Clause and Indian Taxation

and unambiguous provisions of the applicable tax law allowed for a tax benefit to be utilised by the taxpayer. Thus, through a curious set of circumstances, tax law became the battlefield in which the principles of statutory interpretation employed by judges were, in the eyes of the revenue authorities, inadequate to meet the problem of tax avoidance. The government responded to this problem by mandating under the GAAR that a taxpayer arrangement with a tax avoidance motive would be prima facie impermissible under the ITA. The GAAR is ostensibly structured, as described above, as a two stage process, the first of which is the identification of a taxpayer intention to avoid tax and the second stage identifies one of four attributes of avoidance. However, in practice, the two stage process might well be reversed, with the revenue attempting to identify the markers of tax avoidance, and if successful, using the markers to make an argument that the taxpayers entered into their arrangements with the primary purpose of avoiding tax. Therefore the Indian GAAR makes it the case that tax avoidance motivation will now be a factor that the judges have no choice but to consider in interpreting statutes. Why were the judges reluctant in the first place to allow taxpayer intentions to play a role in their interpretations of the tax statutes? To answer this question, we will have to turn to more abstract issues relating to interpretation of tax statutes. The judicial struggle with interpreting tax statutes emanates from the fact that the judges are uncertain about their ability to take into account the intentions of taxpayers in adjudicating on tax motivated transactions. There are two models of interpretations of tax statutes when it comes to addressing tax planning and avoidance measures. The first model is the limited purposive model. Partisans of limited purposive interpretation believe that a purposive interpretation helps the court apply the legislation imaginatively to the facts on hand. When the court is faced with an inquiry into whether a tax statute allows or prohibits a tax benefit, the relevant question to ask if whether the legislature, if it were to be faced with such a situation, can be reasonably thought of as allowing or prohibiting such a tax benefit. Such an inquiry does not limit the judge’s inquiry to the literal meaning of the words of the statute but instead enables the judges to look at the statute’s impact in a commercial and practical context. Therefore, limited purposive interpretation enables the judges to look into the impact of these words in a commercial rather than a dictionary meaning setting. For example, platinum sponge given to employees in lieu of cash can be subject to taxation as if a cash equivalent was provided to employees even though a literal interpretation of the applicable statute will not allow such an application.21 However, unless the statute itself has indicated otherwise, a limited purposive interpretation will not look into taxpayer purposes.22 A judge using a limited purposive interpretation will not take taxpayer intentions into account because that would 21 Hoffmann

(2005), p. 204. legislation is replete with inquiries into taxpayer purposes. For instance, section 37 of the ITA states that a taxpayer can deduct expenses from his business revenues only if the expenditure is wholly and exclusively for the purposes of the taxpayer’ business. Similarly, the distinction between 22 Tax

The Principal Purpose Clause

103

take him beyond the authority of the statute. The idea that judges are acting under the authority of the law is key to an understanding of limited purposive interpretation. The underlying concept of the law at work here is that of a code of conduct emanating from a legislature which is then implemented by the court of law. The code of conduct is authoritative for the judges in the sense that the judges are executing a set of instructions emanating from another source. The instructions might be ambiguous or might contain gaps in which case the judges have to fill in the gaps or decide on what to do with the ambiguity in legal language. However, in filling in gaps or resolving ambiguities, the judges do not go beyond the inquiry about what the legislature was supposed to have provided as a matter of guidance or what the legislature could be taken to have provided as a matter of guidance if it had been appraised of the ambiguities or been aware of the gaps in the laws. Nothing in limited purposive interpretation requires the judge to look into any question extraneous to legislative intention, actual or implied. In contrast with limited purposive interpretation, broad purposive interpretation does not consider the legislation as an authoritative code of conduct that needs to be followed by the judiciary. Broad purposive interpretation considers legislation as purporting to direct the lives of the people who are subject to the politicians that enact the legislation. However, the law that is the outcome of the legislation does not follow, as a matter of linguistics, from the act of legislation. Instead, the outcome of the legislation is a complex process of reasoning that takes into account several considerations emanating from the best moral justification for the legislation. It is the best moral justification for the legislation that is key, not what the legislature believes (or can be taken reasonably to believe) is the purpose of the legislation. Broad purposive interpretation will not be restricted from taking into account taxpayer intentions because it does not purport to bind itself by the language of the statute. What is important is not what the statute is trying to say but what legal position can best exemplify the purpose of the statute. Broad purposive interpretation will allow a judge to take into account taxpayer intentions in allowing or denying a benefit under a tax treaty if the best justification of a tax statute encompasses such an inquiry. Justice Chinnappa Reddy, in McDowell, believed that in a developing country that was short of resources for public welfare, the best justification of tax statutes required that the judiciary deny benefits to what he called as colourable devices. By colourable devices he meant presumably arrangements that were entered into by the taxpayer primarily in order to avoid taxation. Justice Reddy’s judgement in fact was an example of how a broad purposive interpretation can read tax statutes to outlaw tax avoidance purposes. The problem with broad purposive interpretation is that many legal scholars, judges, practitioners and law students believe both that judges do not practice broad purposive interpretation and that judges ought not to practice broad purposive interpretation. While it is not possible for me to defend either of these two propositions comprehensively in this essay, I can give an example for the former and a partial defence of the latter. an asset held as a capital asset or a trading asset depends on the purposes for which the taxpayer holds the asset. Tax law and judges interpreting tax law are familiar with taxpayer purposes.

104

9 BEPS, The Principal Purpose Clause and Indian Taxation

Although the Vodafone case can be considered as an example of either a strict interpretation of tax statutes or a case justifying limited purposive interpretation, it can plausibly be defended as an example of broad purposive interpretation in the Indian tax domain.23 The Vodafone case is perceived by some scholars as the case where judicial benediction was bestowed on a transaction where the primary motive was tax avoidance. However, this is a wrong characterisation of the case. In Vodafone, the Supreme Court Justices found as a matter of fact that the taxpayers had a genuine non tax reason for the arrangements that they entered into-the sale and purchase of a lucrative telecommunications business in India. The purchase was routed through a pre-existing Cayman Islands company in order to avoid Indian taxation but the purchase itself was a genuine transaction. The judges in Vodafone could have reiterated the ambit of section 9 and satisfied themselves with an interpretation that satisfied the linguistic reach of section 9. Instead, the judges went into an elaborate exercise of ascertaining the facts and circumstances of the Hutchinson-Vodafone deal, and decided that the transaction, while motivated by tax considerations, was not primarily motivated by tax considerations.24 This finding was crucial to the decision. The Vodafone decision was premised on the fact that the sale of shares had an overall genuine business purpose behind it even if some of the steps taken to effectuate the transfer had a tax avoidance purposive. One might disagree with the judges’ characterisation of the taxpayer purposes in Vodafone but that is besides the point, which is that the judges did not let the plain wording of the statute prevent them from inquiring into taxpayer purposes. That judges must undertake broad purposive interpretation is a proposition denied by many legal scholars. One common complaint is based on rule of law considerations: since broad purposive interpretation makes the law vague and uncertain, the subjects of the law cannot fathom the requirements of the law before it is applied by the judges and as a consequence the rule of the law is in jeopardy. The problem with this approach is that it gives disproportionate importance to the uncertainties in tax law, which are but common occurrences for tax practitioners at least in the common law countries. Here’s a fairly typical statement from a judge about the distinction between capital and revenue expenditure, a fertile area of litigation in Indian tax law (and elsewhere): It is ultimately a question of law, but a question which must be answered in the light of all the circumstances which are reasonable to take into account, and the weight which must be given to a particular circumstance in a particular case, must depend on common sense rather than on strict application of any single legal principle. It was also observed that solution to the problem is not to be found by any rigid test or description. It has to be derived from many aspects of the whole set of circumstances, some of which may point in one direction, some in the other. One consideration may point so clearly that it dominates other and vaguer

23 (2012)

1 SCR 573. See in particular p. 652 where the court point out that the corporate structure that enabled the Vodafone sale cannot be considered as a ‘tax avoidant’. 24 Ibid.

The Principal Purpose Clause

105

indications in the contrary direction. It is a common sense appreciation of all guiding features which must provide the ultimate answer.25

The judges’ comments on the capital revenue distinction is not atypical. On many matters of tax law, such as the idea of income, capital gains realisation, loan waivers, deemed dividends etc., tax law has been enriched by judicial decisions that were by no means predictable or uncontroversial. Uncertainty is no enemy of the rule of the law, particularly if the taxpayers are able to manage their tax affairs with a fair bit of predictability.

Conclusion: The PPT Rule and GAAR: A Superfluous Solution to a Problem of Statutory Interpretation Both the GAAR and the PP clause mandate, by means of statutory language, that tax authorities must take taxpayer objectives into account in arriving at a conclusion of illegitimate tax avoidance. I believe that taking taxpayer objectives into account is a legitimate move in any effort to combat tax avoidance but it does not need legislation. A broad purposive interpretation of tax legislation would result in taxpayer objectives being taken into account. In fact the introduction of legislative language tends to complicate matters in this domain. Both the PP clause and the GAAR test have additional, supposedly more objective provisions that purport to provide clear criteria for the application of anti avoidance rules. However these so called objective factors ultimately dilute the impact of the PP clause and the GAAR test respectively. The LOB tests that accompany the PP clause place several threshold tests that relate to the ownership of enterprises claiming residency in the treaty countries, or their business activities or the trading of their shares on the stock markets. These markers are objective in the sense that there is neither any need to look into taxpayer intentions nor is there a need to rely on the subjective assessments of revenue authorities. However, the OECD BEPS commentary has made it clear in its commentary on BEPS Action Plan 6 that the fulfillment of the LOB criteria will not immunise the taxpayers from an inquiry under the PP clause.26 The circumstances that the LOB test is trying to address can be addressed under the PP clause; therefore the only reason for the LOB test is to provide greater certainty to the taxpayers. If the LOB test does not act as a hard stop on anti-tax avoidance inquiries under tax treaties and the PP clause can continue to apply to the same set of facts, there continues to be some uncertainty in the application of anti avoidance provisions to taxpayers. Similarly the Indian GAAR provisions have one provision that covers an inquiry into taxpayer purposes whilst at the same time containing another set of provisions that lay out various factors which look into the artificiality of taxpayer arrangements 25 Honda

Siel Cars India Limited v Commissioner of Income Tax, Ghaziabad 2017 Indlaw SC 432, para 21, relying on Alembic Chemical Works Co. Ltd. v. Commissioner of Income Tax, Gujarat (1989) 177 ITR 377, 1989 Indlaw SC 559. 26 BEPS Action Plan 6 Final Report Part A(1)(a) pp. 55–56.

106

9 BEPS, The Principal Purpose Clause and Indian Taxation

and the ability of the arrangements to exploit tax law.27 Once again, it is not clear why these additional factors cannot be assimilated into the first part of the GAAR test which looks at the taxpayer’s purposes in entering into certain arrangements. What the two step process under the Indian GAAR does is to make uncertain the application of anti tax avoidance measures from the point of view of both pragmatics and principle. Pragmatically, the Indian revenue is going to try and prove one of the markers in the second part of the GAAR test and use these arguments to also prove that there was an objective to avoid taxation, thus in effect folding the two part test into one. In terms of principle, one is not sure how most of the markers of tax avoidance in the second part of the GAAR test can be applied without also looking into the intentions of the taxpayer to avoid tax. For example, one is not sure how to assess the artificiality of a taxpayer arrangement or look into the non bona fide aspect of taxpayer relationships without a determination of the nature of taxpayer intentions. If an inquiry into the purposes of the taxpayer is integral to the inquiry into the second part of the GAAR, then one wonders if the elaborate formulation of objective features in the GAAR was necessary. Finally, the relationship between the PPT rule and the GAAR is also likely to lead to uncertainties and complexities in the application of tax avoidance measures. For instance, suppose a foreign institutional investor that invested into India through Mauritius changes tack and makes its investment from a newly acquired corporate structure in Netherlands in order to avoid capital gains in India. However, because of the manner in which the PP clause is framed, there is a very real possibility that the corporate structure will fail the test provided for in the PP clause. The PP clause requires only that one of the primary purposes of a taxpayer arrangement should be to obtain a tax benefit under a tax treaty. Under this test, a taxpayer can have multiple primary purposes and as long as one of the primary purposes is tax avoidance, the PPT rule will capture such an arrangement. The move to Netherlands, even though undertaken ultimately to earn legitimate business profits from India, arguably had tax avoidance as one of its primary motives. The only way out for the taxpayer in such a scenario is to argue that the tax benefit earned by the taxpayer was envisaged as a tax benefit in accordance with the objectives of the treaty. This escape clause might not be available to the taxpayer as he will have to establish that the treaty allows investors to use it as a mere base to invest in other countries without having any significant links with the tax treaty country. It will not be easy to establish such an objective for the treaty in a post BEPS world. To add to the complexity, the corporate structure discussed above might pass the GAAR test. The GAAR test requires that the primary purpose of a taxpayer arrangement must be to avoid taxation. If we construe this provision as requiring one dominant purpose and excluding the possibility of multiple primary purposes, then it is possible for the taxpayer to argue that the primary purpose of moving to Netherlands was to begin a legitimate investment business into India. Tax would be a motivating factor, as it is with any pursuit of a business activity, but it is not a dominant 27 Section

96(1)(a)–(d).

Conclusion: The PPT Rule and GAAR: A Superfluous Solution …

107

objective, even if the only reason for moving from Mauritius to Netherlands was to obtain tax treaty benefits. Given these problems and complexities, what we need is a reassessment of the manner in which tax statutes are interpreted. It is a myth that a statutory anti avoidance rule is necessary to tackle widespread tax evasion. What is needed is for the judiciary to re-discover the art of broad purposive interpretation. However, the process of broad purposive interpretation can be successful only if the international tax community takes a radical look at the judicial settlement of tax disputes. Some efforts towards this idea have begun with the BEPS Action Plans on dispute settlements but the work so far has been patchy and not subscribed to by the majority of the countries. Without an impartial judicial settlement of tax disputes (including disputes about tax avoidance arrangements) it would not be possible for broad purposive interpretation to prosper. Broad purposive interpretation cannot exist as a judicial practice in a domestic or international policy unless there is a concurrent institutional practice of impartial decision making with efficient disposals and appeals. The international tax system must not create a more complicated and intricate system of rules beyond the minimum required for policing the abuse of tax treaties. Towards this end, the introduction of a broad PP clause along with a set of intricate LOB provisions only serves to confuse rather than clarify efforts to combat tax avoidance. It is perhaps too late now to turn the clock back and do away with the PP clause. The second best approach ought to be to retain a simple PP clause and minimum LOB requirements and focus international efforts on improving and encouraging impartial dispute resolution mechanisms in the tax domain, both at the domestic and the international level.

Reference Hoffmann L (2005) ‘Tax Avoidance’ 2005 BTR 197

Conclusion

The international tax system is a complex system of interlocking principles deriving its sustenance from both domestic tax law and a patchwork of bilateral DTAAs. The domestic-international hybrid of rules presents taxpayers with a complex and occasionally impossible challenge of compliance but also provides taxpayers with an opportunity to arrange their commercial affairs in such a way that they can exploit the chinks in the complex armours of international tax rules. Taxpayers, particularly multinational companies took advantage of the inherent flexibility afforded by contractual relationships (sale, purchase, lease, incorporation, restructuring) to minimise their taxes worldwide. Tax law might want to impose a certain taxable status over the commercial activities of companies but the intricate tax planning developed through ingenious contracting by taxpayers has outsmarted frequently the reach of tax legislation. The history of international tax law has so far been a history of the trump of contracts over status. In the international tax arena, the contract v status trope extends to two peculiarities of the system: much tax planning by companies is meant to deny states nexus over their activities, and allied nexus denial is the equally important issue of the manipulation of prices for services and products exchanged between constituent units of the same multinational group (transfer pricing). The issue of nexus is unique to international tax law as each country attempts to extend its territorial nexus over the activities of companies that are not resident in that country. The idea of a source based nexus depends on rules that identify when a non resident person can be said to have a territorial connection to Indian that entitles India to tax the income of the non-resident. When it comes to business income, the Indian rules are subservient to the double tax treaties that India has signed with most of its major trading partners, consequent to which the reigning marker for the taxation of business income in India is that the non-resident person must have a permanent establishment in India. The concept of a permanent establishment has taken on a wide ranging form in Indian tax treaties with various kinds of physical presences (concrete offices, services, dependent agents) recognised as a territorial connection adequate for the assertion of tax jurisdiction by the host country. However, the PE rules have also created opportunities for multinationals to undertake tax planning designed to take © The Author(s), under exclusive licence to Springer Nature India Private Limited 2020 N. Nuggehalli, International Taxation, SpringerBriefs in Law, https://doi.org/10.1007/978-81-322-3670-2

109

110

Conclusion

advantage of the host country resources while at the same time avoiding PE status. Usually this is accomplished through an adroit use of subsidiaries in the host country. The foreign parent will outsource some of its business functions to its Indian subsidiary and achieve significant cost reductions and efficiency because of the lower costs of skilled manpower in India. The subsidiary is usually trained in the business processes through a limited time deputation in India of the foreign parent’s employees. PE rules in Indian tax treaties, following the OECD model, specify explicitly that subsidiaries will not form PEs of their parent companies, merely by virtue of their status as subsidiaries. Further, the PE rules make it clear that only core business activities performed in India will count towards PE status; ancillary or fringe business activities will not count. These two features of Indian PE rules have come together to sanctify outsourcing arrangements of the sort described above. However, the Indian Supreme Court in the famous Morgan Stanley case muddled the PE issue in outsourcing contracts by designating the Indian subsidiary as a service PE of the foreign parent in the event its functions were steered by the foreign parent’s deputationists. However, in the more recent E Funds case, the Supreme Court has clarified that outsourcing functions performed in India would not normally rise to the level of business activity that triggers a PE status in India. While E Funds was favourable to the taxpayer, the Supreme Court in Formula One came down in favour of the revenue in determining if there was a PE in India when the foreign collaborator of an Indian promoter of a car race circuit exercised (contractually) most of the commercial rights associated with the event. However, Formula One was a case in fact of a decision arrived at after an in-depth and comprehensive analysis of the contractual relationship between the Indian promoter and his foreign collaborator. In conclusion, the PE domain provides considerable flexibility for taxpayers to structure their business relationships, with a few decisions like Morgan Stanley bucking the trend by recognising PE status for an Indian subsidiary without providing for a proper reason to do so. In the domain of royalties and technical services, one observes the increasing relevance of status compared to contracts. To begin with, section 9 was amended in 1976 to deem any payment by an Indian resident towards royalty or technical services as Indian source income. A deeming rule of this kind is a departure from the normal territory nexus rules because this rule focused on the residential status of the payor rather than the territorial activities of the payee. In recent years, the territorial nexus principle has been eroded further through statutory amendments to section 9 that broaden the definition of royalty and ensure that royalty and technical payments continue to be sourced in India even if the person receiving such payments has no business presence in India. Similarly, the definition of royalty has changed in as much that the medium through which a right with respect to a copyright is transferred does not matter. Again, in the case of equipment related royalty, it does not matter if the person making use of the equipment is in possession of the equipment. These legislative changes are a glimpse of how the government can attempt to impose status over contract through legislative fiat. While the domains of royalty and technical services have seen legislative changes, the most profound albeit ill fitting legislative changes are to be found in the area

Conclusion

111

of capital gains. The catalyst for the status v contract battle in this area was the famous Vodafone judgment which decreed that non residents could transfer non Indian situs shares between themselves on a tax free basis (because these share transfers did not have an Indian source) even if the shares derived most of their value from underlying Indian property. In response to this decision, the government made a series of sweeping amendments to the definitional provisions in the ITA relating to transfers of property, capital assets, and deemed sources of income. Not only were these amendments wide ranging and open ended, these amendments were also deemed to operate from 1961, thus overturning the decision in the Vodafone case retrospectively. The dominance of status over contract has continued in the area of transfer pricing where for most transactions between associated enterprises, the ITA now requires an arms length price to be determined. Transfer pricing ought to be the poster child for the trump of contract over status because it enables taxpayers to manipulate, through contracts, the pricing of the transactions between them and thus virtually create the extent of their tax liability. The idea of the arms length price, on the other hand ought to be the poster child for the trump of status over contract, as the law imposes a hypothetical price on transactions disregarding the pricing determined by the parties. In practice, multinational enterprises often take advantage of the arms length provisions because comparative pricing figures required to calculate an arms length price are difficult to find and calculate. The Indian revenue has long argued for a formulary apportionment standard for transfer pricing, which allocates the revenue (for tax purposes) of multinationals to countries based on parameters such as the quantum of sales, payroll and assets. However, the formulary apportionment formula has not received much traction worldwide and the arms length standard continues to rule the roost. In India, one of the most contentious areas in transfer pricing has been concerned with AMP (Advertisement, Marketing and Promotion) expenses. The revenue has contended that if an Indian enterprise performs certain brand promotion functions that directly or indirectly promote the brand of its foreign associated enterprise, the Indian enterprise needs to be compensated for conferring a benefit on the foreign enterprise. In practice, this meant that the revenue would challenge any marketing expenses or such enterprises that the revenue deemed as excessive. The courts have addressed this problem by insisting that there must be an agreement between the associated enterprises to promote the brand of one of the associated enterprises. The revenue cannot decide there is a transfer pricing issue by looking merely at the quantum of AMP expenses. Further, the courts have decisively rejected the bright line test (which drew transfer pricing conclusions from the quantum of AMP expenses) as not founded in the Indian statutory regime. The domain of transfer pricing in India has moved towards some contractual freedoms with Advance Pricing Agreements that ensure that as long as prices between associated enterprises are within a pre-determined range, the government will not challenge the prices. The APAs are not as popular as one might have imagined but if they grow in popularity, it will result in a new twist in the contract v status debate,

112

Conclusion

because now the result will be the trump of agreements between the state and the taxpayers over the status of taxpayers determined by the law. The most robust attempt to impose status over contract in the domain of Indian tax law took place with the enactment of GAAR, put into operation in 2017. The GAAR is meant to cure the problems with judicial anti avoidance rules which the government believed took inadequate notice of taxpayer intentions to avoid tax. The GAAR puts taxpayer purposes at the centre of its inquiry by designating certain commercial arrangements as impermissible avoidance arrangements. Impermissible avoidance arrangements are arrangements that have one of its main purposes as the avoidance of tax and in addition, the arrangement has certain features considered as a tax avoidance feature under section 96 of the ITA. In practice, the revenue is likely to begin with the second limb of the GAAR i.e. identify a tax avoidance feature mentioned in section 96 and claim a taxpayer purpose to avoid tax on the basis of such a feature. The second limb of the GAAR involves a tax legislation abuse test and a suspicious characteristics test. While the tax legislation abuse can be accused of being vague, the real problem is the suspicious characteristics test. Are the suspicious characteristics deeming provisions? If so, then a transaction need not be an abuse or a misuse of the tax legislation and yet be a GAAR tainted transaction. Since the GAAR is a relatively new provision, there is no Indian case law jurisprudence yet on the subject, but considering the controversial nature of the provisions, litigation regarding the appropriate application of the GAAR is inevitable in the future. Indian tax legislation is also going to be impacted by the OECD led BEPS project that has culminated in what is called the Multilateral Instrument (MLI). From India’s perspective the most important change is the introduction in its tax treaties of the Principal Purpose (PP) Rule, which uses language similar to the Indian GAAR. Once again, there is no clarity at present on how these two tests will interact with each other in particular situations. As the PP rule becomes more widespread in tax treaties because of the operation of the MLI, its effect vis-à-vis domestic anti tax abuse legislation will be clarified further though model treaty commentaries as well as domestic judicial decisions. The contract v status debate is a confluence of three interlocking principles in tax law: the freedom accorded to persons in determining their commercial affairs contractually, the leeway provided to governments in the area of taxation to expand the reach of the tax system (through the expansion of definitions in section 2 or through the widening of the sourcing rules in section 9) and the ability of statutory interpretation to examine the purposes of commercial arrangements. It is in this battlefield of ideas that the next generation of international tax law principles will be incubated.