Menu
Tax Notes logo

India’s New Profit Attribution Proposal and the Arm’s-Length Standard

Posted on June 17, 2019
Ajitesh Kir
Ajitesh Kir
Reuven S. Avi-Yonah
Reuven S. Avi-Yonah

Reuven S. Avi-Yonah is the Irwin I. Cohn Professor of Law and Ajitesh Kir is an SJD student at the University of Michigan Law School.

A recent Indian public consultation document on amending India’s rules for profit attribution to permanent establishments represents the first time a national government has proposed abandoning the arm’s-length standard. In this article, the authors discuss that document, as well as the OECD’s recent consideration of using formulas to allocate profits to market jurisdictions, and they ponder whether those developments indicate a larger movement away from the arm’s-length standard in favor of a unitary tax system.

On April 18 the Indian Central Board of Direct Taxes released public consultation document F. No. 500/33/2017-FTD.I on amending India’s rules for profit attribution to permanent establishments. That remarkable document is the first time a national government has officially proposed abandoning the arm’s-length standard, which has been the governing standard for dividing business profits among taxing jurisdictions since the 1930s, and replacing it with a system that bears close resemblance to the historic rival of the arm’s-length standard, formulary apportionment.1

Unlike the European Commission’s proposed common consolidated corporate tax base — the other well-known proposal for replacing the arm’s-length standard — the Indian plan has a good chance of becoming law after the national election.2 If it does, it will exert major pressure on the OECD to follow suit and approve formulary alternatives to the arm’s-length standard — as the organization indicated it might be willing to do in its February public consultation document on addressing the tax challenges of the digitalization of the economy.

India’s move might be the first break in the dam of the arm’s-length standard, and even the United States could eventually abandon the arm’s-length standard in favor of a formulary system. That, in turn, could lead to far-reaching changes in the international tax regime, including the abandonment of residence-based corporate taxation in favor of a purely source-based system.

I. The OECD and the Great Digitalization Debate

Action 1 of the OECD’s base erosion and profit-shifting project was supposed to address the tax challenges posed by the digitalization of the economy against the background of unilateral measures like the U.K. diverted profits tax (2015), Australia’s Multinational Anti-Tax-Avoidance Law (2015), and India’s equalization levy (2016).3 However, the OECD was unable to reach consensus on action 1. It kicked the can down the road, releasing an interim report in 2018 and expecting to produce a final report in 2020 to provide a consensus-based long-term solution.

The OECD’s February public consultation document includes two proposals. The first is a global anti-base-erosion minimum tax proposal that builds on the U.S. global intangible low-taxed income regime and the base erosion and antiabuse tax. The idea is to implement what one author has called the “single tax principle” by imposing residence-based taxation when the source tax is too low, and source-based taxation when the residence tax is too low.4 It is a relatively simple extension of existing principles, but it begs the question of how much profit should be allocated to the source jurisdiction. A GILTI-type, residence-based minimum tax works well only if there is a consensus on profit allocation because it envisages granting foreign tax credits for source-country taxes on profits properly allocable to the source country. Similarly, a BEAT-type minimum tax that extends to all deductible payments (including cost of goods sold) requires agreement by the residence jurisdiction to prevent double taxation.5

The OECD recognizes that, and it has laid out three alternative options for profit allocation in the context of digitalization. All three eliminate the physical presence requirement for having a PE but take different approaches to determining how much profits should be subject to tax in the source (market) jurisdiction. The narrowest option is the user participation proposal, which applies only to companies like Facebook, Google, or Amazon and allows the market jurisdiction to tax profits attributable to user participation. It is presumably the EU’s preferred option because it will apply only to U.S. tech giants. The United States countered with a broader option the marketing intangible proposal, which would allocate residuals arising under the traditional profit-split method to the market jurisdiction and would apply to all companies. That option, which the OECD included in its consultation document, builds on the U.S. experience in extracting billions in revenue from EU-based companies based on their use of marketing intangibles in the United States.6

Those two proposals, while radical in abandoning the traditional PE concept, still build on the arm’s-length standard because they allocate residuals to the market (or user) jurisdiction only after routine profits have been allocated using that standard.7

The third OECD option is more extreme, because it not only abandons the PE for a significant economic presence threshold (similar to the proposed EU directive8), but because it also abandons the arm’s-length standard in favor of what it calls a “fractional apportionment method,” as discussed in the action 1 report. That method would require three successive steps: the definition of the tax base to be divided, the determination of the allocation keys to divide that base, and the weighting of those keys. According to the OECD, the tax base could be determined by applying the global profit rate of the multinational group to the revenue generated in a particular jurisdiction and apportioned by taking into account factors such as sales, assets, and employees. For businesses for which users meaningfully contribute to the value-creation process, users would also be considered in apportioning income.

That is a remarkable turnaround, because the OECD has traditionally resisted any attempt to replace the arm’s-length standard with a formulary method.9 The OECD is envisaging applying a multinational entity’s global profit rate to revenue derived in a jurisdiction to determine the potential tax base, and then apportioning that base by using the traditional three-factor formula. It adds users as a factor, because Facebook, for example, might not have any direct sales, assets, or employees in a given jurisdiction.

There is no mention of using the traditional arm’s-length standard when it can be applied and allocating residuals based only on user participation or marketing intangibles, as in the other two proposals. However, the fractional apportionment option is a latecomer and seems unlikely to be favored over the two more traditional and developed options.10 That is why the Indian proposal, which presumably inspired the third OECD option, is so important.

II. The Indian Proposal

India’s proposed amendment packs quite a punch. A carefully drafted 84-page report — replete with economic principles, legal history, international comparisons, and what-have-you — lays out a solid justification for the suggested change. It also indicates that India has possibly lost patience with multilateral decision-making, in which — as it has often argued — capital-exporting countries are disproportionately represented. What began as a litmus test in 2016 with the equalization levy experiment has been amplified by other unilateral decisions, such as the significant economic presence test in 2018.11 But the new proposal regarding profit attribution to a PE is possibly India’s boldest move yet in shaking the international tax order.

What India is proposing is essentially a fractional apportionment method based on profits derived from India. The method differs from formulary apportionment (as followed by the United States and proposed by the EU) by limiting itself to India-related information rather than requiring the consolidation of profits from all relevant tax jurisdictions. It relies primarily on information on Indian operations and adopts a three-factor calculation, according equal weight to sales, employees (manpower and wages), and assets. It also offers a formula for calculating profits derived from India that relies on a nonresident’s India-derived revenue and global operational profit margin. To avoid double taxation of an Indian subsidiary that is also a PE, the proposed method allows a deduction of already taxed income from the apportioned profits. To top it all, the proposal offers a solution to tax the digital economy: It allows user participation to count as the fourth factor in the fractional apportionment method; the relative weight of the user participation may range from 10 to 20 percent, depending on the user intensity of the business enterprise concerned. In sum, the proposed move shuts the door on profit attribution using the function, asset, and risk (FAR) approach and the arm’s-length principle.

Before delving into the details of the profit attribution proposal, let’s take a brief look at the existing structure in India.

Under Indian tax treaties12 and domestic law, profit attribution to a PE adopts the following approach: Once a business enterprise satisfies the taxable threshold nexus of having a business connection in India,13 as well as the taxable threshold nexus of PE, its profits become taxable in India to the extent they are attributable to that country.

The first step, therefore, is to establish the business connection and PE nexus thresholds.

The concept of business connection was substantially amended in 201814 to expand its scope to incorporate recommendations of the final report on action 7 of the BEPS project and include a new threshold of significant economic presence.15 The primary purpose of introducing the significant economic presence rule was to find a way to tax the digital economy.16

According to the 2018 amendment, a nonresident can have significant economic presence in India through two means:

  • if the revenue from its transactions in India (including download of data or software) exceeds a prescribed amount; or

  • if it digitally interacts with a set number of users in India or uses digital means for systematic and continuous solicitation of its business activities.

The significant economic presence threshold creates tax liability in India, regardless of whether the agreement on those transactions was entered into in India, or the nonresident has a place of business in India or renders services there. The significant economic presence rule has yet to be fully tested — the amount of local revenue or number of users required to trigger a significant economic presence have not yet been prescribed. The Indian tax department (as discussed below) wants to first amend the rules of profit attribution to ensure that the significant economic presence threshold works effectively.

For establishing a PE, the requirements in article 5 of the treaties must be satisfied. The definition of PE, which focuses only on physical presence, is wide. It includes concepts such as a fixed place of business, the furnishing of services, construction activities beyond a specific duration, and the role played by a dependent agent.17 In Formula One World Championship Ltd. v. Commissioner of Income Tax, (2017) 15 SCC 602, the central issue before the Supreme Court of India was whether the appellant was a PE under article 5 of the India-U.K. tax treaty. The Court found the appellant to be a PE, noting that a PE must have three characteristics: stability, productivity, and dependence. If a nonresident has a PE in India, then a business connection in India is established.

The second step — that is, profit attribution — is governed by different provisions, depending on whether the tax treaty or the domestic law applies to the nonresident. According to the consultation document, the treaty provisions take precedence over domestic law; a nonresident’s business profits can be taxed under domestic law only if the applicable treaty allows. In either case, only income that is reasonably attributable to the operations carried out in India is subject to tax.18

Indian domestic law has a twofold approach for determining business profits subject to tax. First, an assessing officer looks at the nonresident’s books of accounts. If the books are unavailable, or it is difficult to determine the profits from them, the assessing officer can resort to one of three methods19 (all of which grant the officer wide discretion to calculate profits):

  • at such percentage of the turnover as she considers reasonable (presumptive method);

  • on any amount that bears the same proportion to the total profits as the receipts bear to total business receipts (proportionate method); or

  • in such a manner as she may deem suitable (discretionary method).

The treaty profit attribution principles in article 7 require the business profits attributable to a PE to be what the PE might be expected to make if it were a distinct and separate entity engaged in identical or similar activities under the same or similar conditions independently of the nonresident.20 Article 7(4) contains a caveat, stating that if a jurisdiction customarily attributes a PE’s profits based on an apportionment of the nonresident’s total profits, it may continue to do so.21

In usual practice, when a PE’s separate accounts are available and can be relied on, profit attribution is readily determined and the process is complete. However, when separate accounts are not available or cannot be easily relied on, the Indian tax department invokes article 7(4) to use its domestic law provisions for attribution (in particular, the three methods mentioned above).

Because of the wide discretion in domestic law, there is little uniformity in the choice of method adopted. That has created much uncertainty in tax planning and is a major factor in the rise of litigation on profit attribution.22 Indian courts have approved various methods for attributing profits to a PE, ranging from ad hoc methods to formulary apportionment to activity performed to FAR.

Considering the need for a uniform approach to attributing profits to a nonresident under Indian domestic law, which is routinely used for both treaty and non-treaty cases, India’s proposal purports to usher in clarity and predictability.

The proposal seeks to amend Rule 10 of the Income Tax Rules,23 which addresses the determination of nonresidents’ income. It targets nonresidents who have a business connection in India and derive sales revenue from India through a business whose complete operations are not carried out there. Under the proposal, the income from that kind of business that is attributable to the operations carried out in India (and deemed to accrue or arise there24) is to be apportioned using a formula that equally weights sales, employees, and assets:

Profits attributable to operations in India = Profits derived from India x [SI/3xST + (NI/6xNT) + (WI/6xWT) + (AI/3xAT)]

Where:

SI = sales revenue derived by Indian operations from sales in India;

ST = total sales revenue derived by Indian operations from sales both in and outside India;

NI = number of employees of Indian operations who are located in India;

NT = total number of employees of Indian operations who are located both in and outside India;

WI = wages paid to employees of Indian operations who are located in India;

WT = total wages paid to employees of Indian operations who are located both in and outside India;

AI = assets deployed for Indian operations and located in India; and

AT = total assets deployed for Indian operations and located both in and outside India.

The first component of the above main formula — that is, profits derived from India — is to be calculated using another formula:

Profits derived from India = Revenue derived from India x Global operational profit margin

Where:

  • revenue derived from India includes all receipts arising or accruing (or deemed to arise or accrue) from India; and

  • the margin for earnings before interest, tax, depreciation, and amortization divided by total revenue is taken as the global operational profit margin.

Because that formula will not work — that is, it will produce a negative number — if the global operational profit margin is negative (in case of global operational losses), the proposal contains a backup plan: The profits derived from India would be the higher of either the amount arrived at under the above formula or 2 percent of the revenue derived from India.25 In other words, the profits derived from India — irrespective of global operational losses — cannot be less than 2 percent of the revenue derived from India.

Under the proposal, the formula to determine profits attributable to operations in India will be modified for nonresidents whose business operations have a prescribed user participation in India. That is intended to tax businesses that use digital means to generate revenue, and for which user participation is key.26 It also conforms with the 2018 introduction of the significant economic presence test. Because user participation is a vital component of the total revenue of digital business models, the proposal seeks to incorporate users as a fourth factor in profit attribution. In those cases, the formula would be modified.

For businesses with low- and medium-user intensity, users would be assigned a weight of 10 percent, with the sales, employees, and assets weighing 30 percent each:

Profits attributable to operations in India = Profits derived from India * [0.3 * SI/ST + (0.15 * NI/NT) + (0.15 * WI/WT) + (0.3 * AI/3 * AT)] + 0.1

For businesses with high user intensity, users would be assigned a weight of 20 percent, with employees and assets weighing 25 percent each and sales weighing 30 percent:

Profits attributable to operations in India = Profits derived from India * [0.3 * SI/ST + (0.125 * NI/NT) + (0.125 * WI/WT) + (0.25 AI/3 * AT)] + 0.2

The proposal clarifies that if the nonresident’s PE or business connection in India is an Indian subsidiary, the profits that have already been taxed in India are to be deducted from the profits that are apportioned using the proposed formula. That would prevent double taxation of operations carried out in India through a PE. Also, no further profits will be attributed to the PE if the Indian subsidiary’s receipts for sales or services in India is less than INR 1 million (approximately $14,350), and its activities in India have been fully remunerated by the nonresident through an arm’s-length price.

That method conforms with DIT v. Morgan Stanley, (2007) 7 SCC 1 (concerning the India-U.S. double taxation agreement), in which the Supreme Court of India held that when a nonresident’s associated enterprise in India (that also constitutes a PE) has already been remunerated on an arm’s-length basis and the transfer pricing analysis has been undertaken, no further profits need be attributed to the PE.

It would be wrong to assume that the proposed rule will be used for all profit attribution cases. The proposal makes clear that it would be confined to situations in which the nonresident has not maintained any India-centric financial statements, the books have been rejected under relevant legal provisions of the Companies Act 2013, or, as per the assessing officer, the accounts do not adequately reflect the profits attributable to the PE as an independent and separate entity.

A. Rationale

The economic justification for the proposed fractional method is that because both production and sales — that is, supply and demand — are essential for generating profits, both must be considered for the apportionment thereof. India claims that even though capital-importing countries contribute to demand (mainly by providing a market where consumers are located) and therefore to profit generation, they do not get their fair share of taxing rights over those profits. Thus, the proposal rejects the post-2010 version of OECD model article 7 and the related authorized OECD approach (AOA), arguing that it attributes profits to a PE using only supply-side factors (such as FAR) and ignores demand-side factors (such as location of market, sales, and users). The AOA’s FAR approach to profit attribution, which follows the OECD’s transfer pricing guidelines, denies the market country revenue by artificially reducing its taxable profits.27

The Indian proposal adopts a balanced approach, considering both supply- and demand-side factors in apportioning profits to a PE. It also captures the digital economy, which cannot be properly taxed by the FAR approach, in its four-factor fractional apportionment method.

India’s political justification for the proposal is its rejection of the post-2010 revised article 7 of the OECD model tax convention, as well as the related AOA.28 Instead, Indian tax treaties have consistently adopted article 7 of the U.N. model tax convention, which provides more rights for source-based taxation,29 and the pre-2010 OECD version (with modifications).30 Moreover, India has criticized the revised version for requiring a FAR analysis in the absence of separate accounts for a PE and for omitting the option of apportionment.31 India’s overarching political argument for its unilateral proposal is that the OECD gives precedence to the interests of its members over those of nonmembers.

India’s legal justification is that article 7(4) of both the U.N. model and the pre-2010 OECD model permits the attribution of profits through apportionment, if customary in a state. Because most Indian treaties are based on those versions, the proposed fractional apportionment method is well within the mandate of Indian tax treaties, according to India. It also relies on a March 2018 OECD report on profit attribution that states that tax treaties using a version of article 7 that does not require AOA can adopt a customary domestic law apportionment approach. Moreover, its domestic law (rule 10 of the Income Tax Rules) already allows a kind of profit apportionment. The proposed amendment, which seeks to make the domestic law clearer and more predictable, thus does not appear to be illegal per se.32

B. Possible Sticky Issues

The Indian proposal makes it seem as if the U.N. and pre-2010 OECD versions of article 7 completely omit the arm’s-length principle and implies that transfer pricing concepts were introduced in the field of profit attribution only in the post-2010 OECD version. That is perhaps deliberate to buttress India’s political and legal justifications for the proposal.

A contrary view is that the revised version of article 7 merely made the applicability of the arm’s-length principle more explicit.33 Moreover, the Supreme Court of India has already acknowledged that article 7(2) of the U.N. model advocates the arm’s-length approach for attributing profits to a PE.34

There is concern that the proposed rule could result in double taxation if India’s treaty partners consider it at odds with the treaties, and that the compliance burden for taxpayers might increase.35 Further, businesses with global operational losses might find the minimum 2 percent backup plan unfair.

Even so, the proposed amendment brings much needed clarity and certainty. It also affirms India’s stand as a capital-importing country that wants to move toward a more balanced source-based taxation.36 It is highly likely that the proposal will be passed with minor modifications, if any.

III. The End of the Arm’s-Length Standard?

One observer recently defended the arm’s-length standard and criticized the OECD’s attempt to supplement it with a functional analysis that focuses on where value is created.37 He concluded that the new OECD standard is no better than the fair and reasonable standard that prevailed in the United States before the first regulations under IRC section 482 were promulgated in 1968 and was resurrected in the Ninth Circuit’s new opinion in Altera Corp. v. Commissioner, Nos. 16-70496, 16-70497 (9th Cir. 2019), issued June 7.

He makes a valid point: The OECD BEPS work on transfer pricing has so far been a failure (as the OECD public consultation acknowledges) because it lets multinationals operate in market jurisdictions via a limited risk distributor with limited profits while performing high value-added functions in low-tax jurisdictions. Hence, high-tax jurisdictions have attempted to collect revenue by bypassing tax treaties via non-income taxes (the diverted profits tax, for example).

The OECD public consultation offers two relatively minor fixes: the user participation and marketing intangibles options. The user participation option applies to only a small set of multinationals, and the marketing intangibles option requires a complicated transfer pricing analysis to establish a related residual. The marketing intangibles proposal is not a significant improvement over the current state of affairs and adopting it would merely provide more work for the transfer pricing industry.38

The Indian proposal offers a welcome alternative to the arm’s-length standard. It would be relatively easy to administer and does not require information beyond what is publicly available.

If the OECD were to adopt it, it would likely be adopted all over the world. Even if the OECD balks under U.S. pressure, the Indian proposal might boost the EU’s CCCTB proposal. Even if CCCTB were blocked by Ireland and other likely revenue losers, countries could adopt the Indian proposal unilaterally. Can China be far behind?

In the end, sticking to the arm’s-length standard is rapidly becoming an untenable proposition. If the Indian proposal were to become the norm, there would finally be a viable alternative to the broken arm’s-length standard (and would rectify the mistake the U.S. Congress made in 1962 when it rejected the formula passed by the House in favor of the transfer pricing regulations and subpart F).39

Ideally it makes sense to tax corporations purely at source because corporate residence is relatively meaningless.40 The Indian proposal could make pure source-based taxation a reality and finally bring the international tax regime into the 21st century.

FOOTNOTES

1 On the history of that debate, see Reuven S. Avi-Yonah, “The Rise and Fall of Arm’s Length: A Study in the Evolution of U.S. International Taxation,” 15 Va. Tax Rev. 89 (1995); Avi-Yonah, Kimberly A. Clausing, and Michael C. Durst, “Allocating Business Profits for Tax Purposes: A Proposal to Adopt a Formulary Profit Split,” 9(5) Fla. Tax Rev. 497 (2009); Avi-Yonah, “Between Formulary Apportionment and the OECD Guidelines: A Proposal for Reconciliation,” 2(1) World Tax J. 3 (2010); Avi-Yonah and Ilan Benshalom, “Formulary Apportionment — Myths and Prospects,” 3(3) World Tax J. 371 (2011); and Sol Piciotto, Taxing Multinational Enterprises as Unitary Firms (2017).

2 The election has brought in the same ruling alliance in power. The ruling party (BJP) won 303 seats, taking the total tally of the ruling alliance (NDA) to 353 seats in the 17th Lok Sabha (Lower House of Parliament). Elections were held for 543 seats. See Election Commission of India, “Election Commission of India Submits List of Newly Elected Members to the 17th Lok Sabha to the Hon’ble President of India,” Press Note No. ECI/PN/60/2019 (May 25, 2019).

3 For discussion, see Avi-Yonah and Haiyan Xu, “Evaluating BEPS: A Reconsideration of the Benefits Principle and Proposal for UN Oversight,” 6(2) Harv. Bus. L. Rev. 185 (2016); Avi-Yonah, “Three Steps Forward, One Step Back? Reflections on ‘Google Taxes’ and the Destination-Based Corporate Tax,” 2 Nordic Tax J. 1 (2016).

4 See Avi-Yonah, “International Taxation of Electronic Commerce,” 52(3) Tax L. Rev. 507 (1997); Avi-Yonah, “Who Invented the Single Tax Principle? An Essay on the History of US Treaty Policy,” 59(2) NYLS L. Rev. 305 (2015); Avi-Yonah and Gianluca Mazzoni, “Complete Distributive Rules and the Single Tax Principle: A Review of Recent Italian Case Law,” 73(4) Bull. for Int’l Tax’n (Feb. 20, 2019).

5 That problem is why the BEAT did not retain the U.S. House proposal to include cost of goods sold, even though omitting them leaves a gaping hole.

6 Take the GlaxoSmithKline case, for example. It involved an asserted deficiency of $30 billion as a result of the IRS’s contention that more profit from Zantac, a drug developed in the United Kingdom and sold in the United States, should be allocated to marketing intangibles. The case was settled in 2006 for $3.4 billion, but the United Kingdom refused to accept the shift of $10 billion in profit to the United States, so Glaxo was double taxed. See IR-2006-142. See also Reuven S. Avi-Yonah and Gianluca Mazzoni, “Taking the First Bite: Who Should Tax Apple’s $187 Billion in Ireland?University of Michigan Law & Economics Research Paper No. 16-033 (Feb. 22, 2017).

7 That is similar to a recent academic proposal: See Michael P. Devereux et al., “Residual Profit Allocation by Income,” Oxford University Centre for Business Taxation WP19/01 and Working Paper of the Max Planck Institute for Tax Law and Public Finance No. 2019-04 (Mar. 22, 2019).

8 See European Parliament, “Corporate Taxation of a Significant Digital Presence,” Briefing (2018).

10 The OECD noted that the discussion has focused primarily on the user participation and marketing intangible proposals, adding that “a detailed discussion of the concept of significant economic presence is also taking place, but this concept was revisited more recently.”

11 India introduced that concept under the Finance Act 2018, thereby expanding the scope of taxation of nonresidents there. It is intended to tax business models operating remotely through digital means without having any physical presence in India, resulting in avoidance of taxation. For more details, see explanation 2A to section 9(1)(i) of Income Tax Act, 1961.

12 Indian tax treaties largely follow the U.N. model tax convention (with variations depending on the treaty partner). India has a tax treaty network with 96 countries. See Indian Income Tax Department, Double Taxation Avoidance Agreements (undated).

13 See ITA section 9(1)(i).

14 Finance Act, 2018, section 4. The amendment took effect April 1, 2019.

15 See explanations 2 and 2A to ITA section 9(1)(i).

16 Vishal Anand and Arjun Khandelwal, “Decoding India’s ‘Significant Economic Presence,’” Taxsutra (Aug. 13, 2018); and Suranjali Tandon, “India — Tax Challenges Arising From Digitalization,” 24(6) Asia-Pacific Tax Bulletin (2018).

17 See, e.g., the amended India-Mauritius treaty (Aug. 10, 2016).

18 See explanation 1 to ITA section 9(1)(i).

19 Income Tax Rules, 1962, rule 10; consultation document, at 11.

20 That embodies the arm’s-length principle. The Indian position is that article 7(2) of its treaties is based on the U.N. and pre-2010 OECD model conventions, which leave room for attributing profits using a PE’s separate accounts. The post-2010 OECD version, which India has rejected, attributes profits using functions performed, assets used, and risks assumed.

21 The India-Mauritius treaty has a slight variation: Article 7(4) does not mention any customary right to apportionment. Article 7(2) provides that the profit attribution may be estimated on a reasonable basis if it cannot be readily determined or is exceptionally difficult to determine.

22 See, e.g., ZTE Corp. v. ADIT, (2016) 159 ITD 696 (Del); CIT v. Hyundai Heavy Industries Co. Ltd., (2007) 291 ITR 482 (SC); Rolls Royce PLC v. DIT International Taxation, (2011) 339 ITR 147 (Del); Hukum Chand Mills Ltd. v. CIT, (1976) 103 ITR 548; and Arrow Electronics India Ltd. v. Additional Director of Income Tax, IT (TP)A.209 and 210, Bang/2011 (Mar. 31, 2017).

23 That would be an amendment to the central government’s rules and regulations, not the main statute. See ITA section 295.

24 ITA section 9(1)(i).

25 The supposed rationale behind that backup is that even if the nonresident’s Indian operations are indeed profitable, the formula will fail if the nonresident’s global operations have operational losses.

26 See OECD 2018 interim report; and U.K. Treasury, “Corporate Tax and the Digital Economy: Position Paper & Update” (Mar. 13, 2018).

27 Carlos Gutiérrez, “The UN Model and the BRICS Countries — Another View,” in Taxation of Business Profits in the 21st Century — Selected Issues Under Tax Treaties (2013).

29 Klaus Vogel, “Worldwide vs. Source Taxation of Income — A Review and Re-Evaluation of Arguments,” 16(8/9) Intertax 216 (1988).

30 The main differences between the U.N. and the pre-2010 OECD model conventions are that the U.N. version contains a force of attraction clause and prohibits deductions for some PE payments to its head office.

31 OECD, “Positions on Article 7 and Its Commentary” (Nov. 21, 2017), at 625.

32 India also argues that various Indian courts have accepted a type of fractional apportionment as a permissible method.

33 Rahul Mitra, “CBDT Draft Guidelines on PE Profit Attribution — Not at Arm’s Length With Treaty Provisions?” Taxsutra (Apr. 22, 2019).

34 See Morgan Stanley, (2007) 7 SCC 1, at para. 34.

35 EY, “India Tax Alert — News from Transfer Pricing” (Apr. 22, 2019).

36 Avi-Yonah, “Between Formulary Apportionment and the OECD Guidelines: A Proposal for Reconciliation,” University of Michigan Law School, Law & Economics Working Paper (2009).

37 David L. Forst, “One World, Two Transfer Pricing Laws: What’s to Be Done?Tax Notes, Apr. 22, 2019, p. 551.

38 As the Glaxo case shows. See note 6, supra.

39 See note 1, supra. The 1930 League of Nations’ draft, which is the ancestor of modern treaties and included the arm’s-length standard, envisaged formulary apportionment as an alternative for the allocation of profits to PEs when the arm’s-length standard did not work. See Philip Baker, “The League of Nations’ Draft Convention for the Allocation of Business Income Between States — A New Starting Point for the Attribution of Profits to Permanent Establishments,” 5 Brit. Tax Rev. 514 (2018).

40 Avi-Yonah, “Slicing the Shadow: A Proposal for Updating U.S. International Taxation,” Tax Notes, Mar. 15, 1993, p. 1511.

END FOOTNOTES

Copy RID