Canada v. Alta Energy Luxembourg S.A.R.L., 2021 SCC 49, [2021] 3 SCR 590

SUPREME COURT OF CANADA
Citation: Canada v. Alta Energy Luxembourg S.A.R.L., 2021 SCC 49, [2021] 3 S.C.R. 590			Appeal Heard: March 19, 2021 Judgment Rendered: November 26, 2021 Docket: 39113
Between:   Her Majesty The Queen Appellant   and   Alta Energy Luxembourg S.A.R.L. Respondent     Coram: Wagner C.J. and Abella, Moldaver, Karakatsanis, Côté, Brown, Rowe, Martin and Kasirer JJ.
Reasons For Judgment: (paras. 1 to 97)	Côté J. (Abella, Moldaver, Karakatsanis, Brown and Kasirer JJ. concurring)
Joint Dissenting Reasons: (paras. 98 to 189)	Rowe and Martin JJ. (Wagner C.J. concurring)

 

 

 

Her Majesty The Queen                                                                                 Appellant

v.

Alta Energy Luxembourg S.A.R.L.                                                           Respondent

Indexed as: Canada v. Alta Energy Luxembourg S.A.R.L.

2021 SCC 49

File No.: 39113.

2021: March 19; 2021: November 26.

Present: Wagner C.J. and Abella, Moldaver, Karakatsanis, Côté, Brown, Rowe, Martin and Kasirer JJ.

on appeal from the federal court of appeal

                    Taxation — Income tax — Tax avoidance — Application of general anti‑avoidance rule — Large capital gain realized by corporate resident of Luxembourg on sale of shares whose value derived principally from immovable property situated in Canada — Corporation claiming exemption from Canadian tax on basis that shares were protected property under tax treaty between Canada and Luxembourg — Whether general anti‑avoidance rule applicable to deny requested exemption — Income Tax Act, R.S.C. 1985, c. 1 (5th Supp.), s. 245 — Convention between the Government of Canada and the Government of the Grand Duchy of Luxembourg for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and on Capital, Can. T.S. 2000 No. 22, art. 13.

                    In 2011, two American firms founded an American company for the purpose of acquiring and developing unconventional oil and natural gas properties. Alta Energy Partners Canada Ltd. (“Alta Canada”), a wholly owned Canadian subsidiary of that company, was incorporated in order to carry on that business. A restructuring of Alta Canada was undertaken in 2012. As part of the restructuring, Alta Energy Luxembourg S.A.R.L. (“Alta Luxembourg”) was incorporated under the laws of Luxembourg and its shares were issued to a new Canadian partnership. On the same day, Alta Luxembourg purchased all of the shares of Alta Canada. In 2013, it sold those shares, realizing a capital gain in excess of $380 million. Payment for the shares was organized so that Alta Luxembourg did not receive any of the sale proceeds. Following the sale, Alta Luxembourg did not conduct any other business or hold any other investments.

                    The capital gain was reported to the Luxembourg tax authorities and was subject to full taxation under Luxembourg’s domestic laws. In its Canadian tax return for 2013, Alta Luxembourg claimed an exemption from Canadian tax on the basis that the gain was not included in its “taxable income earned in Canada” under s. 115(1)(b) of the Income Tax Act (“Act”) because the shares were “treaty‑protected property” under art. 13(4) and (5) of the Convention between the Government of Canada and the Government of the Grand Duchy of Luxembourg for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and on Capital (“Treaty”). Article 13(4) of the Treaty creates an exemption for residents of Luxembourg from Canadian tax arising from a capital gain on the alienation of shares the value of which is derived principally from immovable property situated in Canada and in which the business of the company was carried on.

                    The Minister denied the treaty exemption. Alta Luxembourg appealed to the Tax Court of Canada. The Minister argued that the business property exemption in art. 13(4) of the Treaty did not apply and, in the alternative, if the shares did qualify as treaty‑protected property, that the general anti‑avoidance rule (“GAAR”) in s. 245 of the Act should apply. The Tax Court found that the shares were treaty‑protected property. With respect to the GAAR, the parties agreed that the restructuring was an “avoidance transaction” as defined in s. 245(3) of the Act that resulted in a tax benefit. The Tax Court held that the avoidance transaction did not result in a misuse or abuse of the provisions of the Act or the Treaty. The Federal Court of Appeal dismissed the Minister’s appeal, which raised only the issue of whether the GAAR applied.

                    Held (Wagner C.J. and Rowe and Martin JJ. dissenting): The appeal should be dismissed.

                    Per Abella, Moldaver, Karakatsanis, Côté, Brown and Kasirer JJ.: The Minister has not discharged her burden of proving abusive tax avoidance. In agreeing to include a specific exemption for immovable property in the Treaty, Canada sought to encourage investments by Luxembourg residents in business assets embodied in immovable property located in Canada and to reap the ensuing benefits. Alta Luxembourg made exactly such an investment. It is a resident of Luxembourg and, as such, is exempt from Canadian taxes on the capital gain realized on the disposition of shares of its wholly owned Canadian subsidiary.

                    The GAAR acts as a legislative limit on both tax certainty and the well‑accepted principle that taxpayers are entitled to arrange their affairs to minimize the amount of tax payable. It bars abusive tax avoidance transactions, including those in which taxpayers seek to obtain treaty benefits that were never intended by the contracting states, but it cannot be used to fundamentally alter the criteria under which a person is entitled to the benefits of a treaty. Applying the GAAR involves a three‑part process meant to determine: (1) whether there is a tax benefit arising from a transaction; (2) whether the transaction is an avoidance transaction; and (3) whether the avoidance transaction is abusive. To determine whether a transaction is abusive, the Court has set out a two‑step inquiry. Under the first step, the provisions relied on for the tax benefit are interpreted to determine their object, spirit, and purpose. In cases of treaty interpretation, this must be done with a view to implementing the true intention of the parties. Under the second step, a factual analysis determines whether the avoidance transaction at issue frustrates the object, spirit, and purpose of the provisions.

                    The object, spirit, and purpose of the business property exemption provided for in art. 13(4) and (5) of the Treaty are to foster international investment. The object, spirit, and purpose of arts. 1 and 4, which make residence central to the application of the Treaty, are to allow all persons who are residents under the laws of one or both of the contracting states to claim benefits under the Treaty, so long as their resident status could expose them to full tax liability. According to art. 4, “residence” under the Treaty is based on liability to tax in one or both of the contracting states by reason of domicile, residence, place of management or another similar criterion. In the context of corporations, the “liable to tax” requirement is met where the domestic law of a contracting state exposes a corporation to full tax liability because it has its residence in that state. Residence is to be defined by the laws of the contracting state in which residence is claimed. Consistent with international practice, Luxembourg law grants resident status to corporations having either their legal seat or their central management in Luxembourg. This does not depart from accepted usage such that the bargain struck in the Treaty could be upheld only if Luxembourg residents claiming benefits have sufficient substantive economic connections to their country of residence. If the drafters had truly intended to include only corporations with sufficient substantive economic connections to their country of residence within the scope of the Treaty, they would have clearly signalled their intention to depart from a well‑established criterion like the “place of incorporation” or “legal seat”.

                    Another contextual element further reinforces the conclusion that the purpose of arts. 1 and 4 of the Treaty is not to reserve benefits to corporations with sufficient substantive economic connections to their country of residence: the inclusion of art. 28(3) in the Treaty, which denies benefits to certain Luxembourg holding companies. The parties’ choice of this approach should be understood as a rejection of the relevance of economic ties for delineating which corporations should be entitled to benefits and which should not. This choice suggests that the drafters intended to exclude a corporation with minimal economic connections to one of the contracting states only where the corporation is a holding company benefiting from Luxembourg’s well known international tax haven regime. In light of this clear intention, the spirit of arts. 1 and 4 was not to limit access to the benefits of the Treaty to corporations with sufficient substantive economic connections to their country of residence.

                    Although the absence of specific anti‑avoidance rules is not necessarily determinative of the application of the GAAR, their absence sheds light on the contracting states’ intention. This is not a case where Parliament did not or could not have foreseen the tax strategy employed by the taxpayer. The use of conduit corporations — legal entities created in a state essentially to obtain treaty benefits that would not be available directly — was not an unforeseen tax strategy at the time of the Treaty. Options to remediate the situation were available and known by the parties, but they made deliberate choices to guard some benefits against conduit corporations and to leave others unguarded. Had the parties truly intended to prevent such corporations from taking advantage of the business property exemption, they could have done so. Combined with Canada’s preference at the time of the Treaty for taking advantage of the economic benefits yielded by foreign investments rather than higher tax revenues, this makes the rationale of the business property exemption even clearer. The fact that the capital gains may not be taxed in Luxembourg, leading to double non‑taxation, and the fact that conduit corporations can take advantage of the business property exemption are tax planning outcomes consistent with the bargain struck between Canada and Luxembourg.

                    In raising the GAAR, Canada is now seeking to revisit its bargain in order to secure both foreign investments and tax revenues. Tax treaties are replete with choices. One key choice made by Canada and Luxembourg in negotiating the Treaty was to deviate from the OECD Model Tax Convention on Income and on Capital by allocating to a person’s residence state the right to tax capital gains realized on the disposition of shares or other similar interests deriving their value principally from immovable property used in a corporation’s business. The business property exemption is a clear departure from the theory of economic allegiance, under which the parties to a treaty avoid double taxation by allocating the right to collect taxes to the contracting state to which the income and the taxpayer are more closely connected. Canada effectively agreed to give up its right to tax certain entities incorporated in Luxembourg in exchange for the jobs and economic opportunities that the business property exemption would promote.

                    The provisions of the Treaty operated as they were intended to operate —the avoidance transaction neither defeated nor frustrated the object, spirit, or purpose of the provisions in issue. Therefore, there was no abuse, so the GAAR cannot be applied to deny the tax benefit claimed. The Treaty makes it clear that Canada and Luxembourg agreed that the power to tax would be allocated to Luxembourg where the conditions of the business property exemption were met. There is nothing in the Treaty suggesting that a single‑purpose conduit corporation resident in Luxembourg cannot avail itself of the benefits of the Treaty due to some other consideration. The provisions of the Treaty operated as they were intended to operate; there was no abuse, and, therefore, the GAAR cannot be applied to deny the tax benefit claimed.

                    Per Wagner C.J. and Rowe and Martin JJ. (dissenting): The appeal should be allowed. Alta Luxembourg’s claim for a tax benefit under the Treaty is the result of abusive avoidance transactions. The courts below did not properly identify the rationale underlying the relevant provisions of the Treaty. They gave weight only to the text and failed to consider why the provisions were put in place. This is not the exercise mandated under the GAAR. Technical compliance with a tax treaty in a way that frustrates the underlying rationale of the provisions relied upon by the taxpayer is precisely what triggers the GAAR. Although it is a long standing principle in Canadian law that taxpayers may arrange their affairs to minimize their amount of tax payable, an unbridled application of that principle can mislead taxpayers into believing that tax plans that merely comply with the technical provisions of the Act are acceptable. Similarly, treaty shopping is not inherently abusive, but where taxing rights in a tax treaty are allocated on the basis of economic allegiance and conduit entities claim tax benefits despite the absence of any genuine economic connection with the state of residence, treaty shopping is abusive.

                    Canada has acted to curb abusive international tax avoidance by enacting the GAAR, which denies tax benefits when taxpayers engage in transactions that conform with the text of the tax rules relied upon, but do not accord with their rationale. As such, the GAAR vests upon courts the unusual duty to look beyond the words of the applicable provisions to determine whether the transactions in question frustrate their underlying rationale. An interpretation confined to the black letter of these legislative provisions would defeat Parliament’s will and fail to fulfil the courts’ role. As the question under the GAAR is not whether the taxpayer can claim a tax benefit, but rather why the benefit was conferred, a GAAR analysis is not constrained by the text in the same way as a traditional statutory interpretation. While this gives rise to a degree of uncertainty for taxpayers, allowing the GAAR to create this uncertainty was a deliberate choice that Parliament made when it enacted a provision that can defeat tax avoidance schemes that exploit Canada’s legislation and treaties. Where those schemes cross the line into abusive tax avoidance, a finding that the GAAR applies does not run counter to the principles of certainty, predictability and fairness.

                    The allocation of taxing powers in the Treaty follows the theory of “economic allegiance”, so the object, spirit or purpose of the relevant provisions of the Treaty is to assign taxing rights to the state with the closest economic connection to the taxpayer’s income. Under art. 13(5), the state of residence retains its jurisdiction to tax capital gains unless the exceptions in art. 13(1) to (4) apply. Article 13(1) preserves the right of the source state to tax gains derived from immovable property situated in that state, and art. 13(4) preserves the source state’s right to tax capital gains arising from the disposition of shares the value of which is derived principally from immovable property situated in that state, unless the company carries on business in the property. The business property exemption assigns the right to tax capital gains arising from the disposition of immovable property in which business is carried on to the resident state. The rationale behind the business property exemption is to encourage investment; it reflects the fact that the business activity, rather than the immovable property itself, drives the value of the property. Article 13(4) therefore allocates to Luxembourg the right to tax its residents’ indirect gains from immovable property situated in Canada used in a business.

                    In the instant case, the abuse is clear. Alta Luxembourg had no genuine economic connections with Luxembourg as it was a mere conduit interposed in Luxembourg for residents of third‑party states to avail themselves of a tax exemption under the Treaty. This lack of any genuine economic connection to Luxembourg frustrates the rationale of the relevant provisions of the Treaty. The federal government did not deliberately set out to create the conditions for unlimited tax avoidance by means of schemes such as that in which Alta Luxembourg was used. The Court should not legitimize such blatantly abusive tax avoidance based on the view that Canada should have negotiated different treaty terms. Ex ante speculation about how the treaty parties ought to have proceeded based on alternatives said to have been available to them gives primacy to what is not there. Parliament was entitled to rely on the GAAR to address abusive uses of the Treaty rather than negotiate the inclusion of a specific rule. The focus should be on what was actually agreed upon and whether the underlying rationale of the relevant provisions was frustrated by the avoidance transactions undertaken. In the give and take of treaty negotiation, Canada certainly did not give up the GAAR.

                    The facts of this case are a patent example of a sophisticated taxpayer effecting a restructuring on the basis of professional tax advice to avoid Canadian tax. In such cases, the principle of fairness ought not to be ignored. As for the degree of uncertainty introduced by the GAAR, it is counterbalanced by the Crown’s burden to show that the avoidance transactions frustrates the object, spirit or purpose of the provisions relied on by the taxpayer and by the fact that any doubt under the GAAR analysis is to be resolved in favour of the taxpayer.

Cases Cited

By Côté J.

                    Applied: Canada Trustco Mortgage Co. v. Canada, 2005 SCC 54, [2005] 2 S.C.R. 601; Copthorne Holdings Ltd. v. Canada, 2011 SCC 63, [2011] 3 S.C.R. 721; approved: Prévost Car Inc. v. Canada, 2009 FCA 57, [2010] 2 F.C.R. 65; referred to: R. v. MIL (Investments) S.A., 2007 FCA 236, [2007] 4 C.T.C. 235; MIL (Investments) S.A. v. R., 2006 TCC 460, [2006] 5 C.T.C. 2552; Commissioners of Inland Revenue v. Duke of Westminster, [1936] A.C. 1; Lipson v. Canada, 2009 SCC 1, [2009] 1 S.C.R. 3; R. v. Melford Developments Inc., [1982] 2 S.C.R. 504; Crown Forest Industries Ltd. v. Canada, [1995] 2 S.C.R. 802; Stubart Investments Ltd. v. The Queen, [1984] 1 S.C.R. 536; J. N. Gladden Estate v. The Queen, [1985] 1 C.T.C. 163; Cudd Pressure Control Inc. v. R., [1999] 1 C.T.C. 1; SA Andritz, No. 233894, Conseil d’État (Section du Contentieux), December 30, 2003; R. v. Hauser, [1979] 1 S.C.R. 984; R. v. McLeod (1950), 97 C.C.C. 366; Dilworth v. Commissioner of Stamps, [1899] A.C. 99; Shell Canada Ltd. v. Canada, [1999] 3 S.C.R. 622; Sun Indalex Finance, LLC v. United Steelworkers, 2013 SCC 6, [2013] 1 S.C.R. 271.

By Rowe and Martin JJ. (dissenting)

                    Commissioners of Inland Revenue v. Duke of Westminster, [1936] A.C. 1; Canada Trustco Mortgage Co. v. Canada, 2005 SCC 54, [2005] 2 S.C.R. 601; Shell Canada Ltd. v. Canada, [1999] 3 S.C.R. 622; Copthorne Holdings Ltd. v. Canada, 2011 SCC 63, [2011] 3 S.C.R. 721; Lipson v. Canada, 2009 SCC 1, [2009] 1 S.C.R. 3; RMM Canadian Enterprises Inc. v. R., [1998] 1 C.T.C. 2300; McNichol v. The Queen, 97 D.T.C. 111; MIL (Investments) S.A. v. R., 2006 TCC 460, [2006] 5 C.T.C. 2552; Prévost Car Inc. v. Canada, 2009 FCA 57, [2010] 2 F.C.R. 65; Crown Forest Industries Ltd. v. Canada, [1995] 2 S.C.R. 802; Fundy Settlement v. Canada, 2010 FCA 309, [2012] 2 F.C.R. 374; R. v. MIL (Investments) S.A., 2007 FCA 236, [2007] 4 C.T.C. 235; Fundy Settlement v. Canada, 2012 SCC 14, [2012] 1 S.C.R. 520; Canada v. 594710 British Columbia Ltd., 2018 FCA 166, [2019] 5 C.T.C. 1.

Statutes and Regulations Cited

Budget Implementation Act, 2004, No. 2, S.C. 2005, c. 19, s. 52(2).

Canada‑United States Tax Convention Act, 1984, S.C. 1984, c. 20, Sch. I [the Convention Between Canada and the United States of America with Respect to Taxes on Income and on Capital].

Income Tax Act, R.S.C. 1985, c. 1 (5th Supp.), ss. 2(3), 38, 39, 66(15) “principal‑business corporation”, 115(1)(b), 245 [repl. 2005, c. 19, s. 52], 248(1), 250(4)(a).

Income Tax Conventions Implementation Act, 1999, S.C. 2000, c. 11, s. 51(2).

Income Tax Conventions Interpretation Act, R.S.C. 1985, c. I‑4, s. 4.1.

Multilateral Instrument in Respect of Tax Conventions Act, S.C. 2019, c. 12.

Treaties and Other International Instruments

Agreement between the Government of Canada and the Government of the Federal Republic of Nigeria for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and on Capital Gains, Can. T.S. 1999 No. 48, arts. 10(7), 11(8), 12(7).

Convention between the Government of Canada and the Government of the Grand Duchy of Luxembourg for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and on Capital, Can. T.S. 2000 No. 22, arts. 1, 3, 4, 5, 6(2), 10 to 12, 13, 28(3).

Convention between the Government of Canada and the Government of the Republic of Kazakhstan for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and on Capital, Can. T.S. 1998 No. 13, arts. 11(8), 12(7), 28(3).

Convention between the Government of Canada and the Government of the Republic of Peru for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and on Capital, Can. T.S. 2002 No. 23, arts. 10(7), 11(7), 12(7), 28(3).

Convention between the Government of Canada and the Government of the Republic of Uzbekistan for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and on Capital, Can. T.S. 2000 No. 18, arts. 11(8), 12(7).

Convention between the Government of Canada and the Government of Ukraine for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and on Capital, Can. T.S. 1997 No. 39, arts. 11(8), 12(8).

Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting, Can. T.S. 2019 No. 26.

Vienna Convention on the Law of Treaties, Can. T.S. 1980 No. 37, arts. 26, 31.

Authors Cited

Anonymous. “The big carve‑up”, The Economist, May 15, 2021, pp. 65‑66.

Arnold, Brian J. “Reflections on the Relationship Between Statutory Interpretation and Tax Avoidance” (2001), 49 Can. Tax J. 1.

Arnold, Brian J. Reforming Canada’s International Tax System: Toward Coherence and Simplicity. Toronto: Canadian Tax Foundation, 2009.

Arnold, Brian J. “Tax Treaties and Tax Avoidance: The 2003 Revisions to the Commentary to the OECD Model” (2004), 58 I.B.F.D. Bulletin 244.

Arnold, Brian J., and James R. Wilson. “Aggressive International Tax Planning by Multinational Corporations: The Canadian Context and Possible Responses”, in University of Calgary School of Public Policy, Vol. 7, Research Paper No. 29, 2014.

Avi‑Yonah, Reuven S., Nicola Sartori and Omri Marian. Global Perspectives on Income Taxation Law. New York: Oxford University Press, 2011.

Baker, Philip. The United Nations Model Double Taxation Convention Between Developed and Developing Countries: Possible Extension of the Beneficial Ownership Concept, U.N. Doc. E/C.18/2008/CRP.2/Add.1, Ann., October 17, 2008.

Bammens, Niels, and Luc De Broe. “Treaty Shopping and Avoidance of Abuse”, in Michael Lang et al., eds., Tax Treaties: Building Bridges between Law and Economics. Amsterdam: IBFD, 2010, 51.

Baruch, Jérémie, et autres. OpenLux: Enquête sur le Luxembourg, coffre-fort de l’Europe, 8 février 2021 (online: https://www.lemonde.fr/les-decodeurs/visuel/2021/02/08/openlux-enquete-sur-le-luxembourg-coffre-fort-de-l-europe_6069132_4355770.html).

Brown, Catherine Anne, and Joseph Bogle. “Treaty Shopping and the New Multilateral Tax Agreement — Is it Business as Usual in Canada?” (2020), 43 Dal. L.J. 1.

Canada. Report of the Royal Commission on Taxation, vol. 3, part A, Taxation of Income. Ottawa, 1966.

Canada. Department of Finance. Consultation Paper on Treaty Shopping — The Problem and Possible Solutions, August 12, 2013 (online: https://web.archive.org/web/20140304022356/http://www.fin.gc.ca/activty/consult/ts-cf-eng.asp; archived version: https://www.scc-csc.ca/cso-dce/2021SCC-CSC49_1_eng.pdf).

Canada. Department of Finance. Explanatory Notes to Legislation Relating to Income Tax. Ottawa, 1988.

Canada. Office of the Auditor General. Report of the Auditor General of Canada to the House of Commons — Chapter 7 — Canada Customs and Revenue Agency, International Tax Administration: Non‑Residents Subject to Canadian Income Tax. Ottawa, 2001.

Canada. Office of the Auditor General. Report of the Auditor General of Canada to the House of Commons, 1992. Ottawa, 1992.

Canada. Senate. Standing Senate Committee on National Finance. Proceedings of the Standing Senate Committee on National Finance, No. 19, 1st Sess., 38th Parl., April 20, 2005.

Canada. Senate. Standing Senate Committee on National Finance. Proceedings of the Standing Senate Committee on National Finance, No. 20, 1st Sess., 38th Parl., May 2, 2005.

Christians, Allison, and Nicolas Benoît‑Guay. “Status and Structure of Tax Treaties”, in Jean‑Pierre Vidal, ed., Introduction to International Tax in Canada, 9th ed. Toronto: Thomson Reuters, 2021.

Couzin, Robert. Corporate Residence and International Taxation. Amsterdam: IBFD, 2002.

Davies, David R. Principles of International Double Taxation Relief. London: Sweet & Maxwell, 1985.

Diamond, Walter H., et al. Tax Havens of the World, vol. 2. New York: Matthew Bender, 2020 (loose‑leaf updated 2020, release 170‑3).

Dodge, David A. “A New and More Coherent Approach to Tax Avoidance” (1988), 36 Can. Tax J. 1.

Duff, David G. “Tax Treaty Abuse and the Principal Purpose Test — Part 1” (2018), 66 Can. Tax J. 619.

Duff, David G. “The Interpretive Exercise Under the General Anti-Avoidance Rule”, in Brian J. Arnold, ed., The General Anti‑Avoidance Rule — Past, Present, and Future. Toronto: Canadian Tax Foundation, 2021, 383.

Grundy’s Tax Havens: A World Survey, 4th ed., by John Walters, ed. London: Sweet & Maxwell, 1983.

Hattingh, P. Johann. “Article 1 of the OECD Model: Historical Background and the Issues Surrounding It” (2003), 57 Bull. Int’l Fisc. Doc. 215.

Innes, William I., Patrick J. Boyle and Joel A. Nitikman. The Essential GAAR Manual: Policies, Principles and Procedures. Toronto: CCH Canadian, 2006.

Kemmeren, Eric C. C. M. “Legal and Economic Principles Support an Origin and Import Neutrality‑Based over a Residence and Export Neutrality‑Based Tax Treaty Policy”, in Michael Lang et al., eds., Tax Treaties: Building Bridges between Law and Economics. Amsterdam: IBFD, 2010, 237.

Knechtle, Arnold A. Basic Problems in International Fiscal Law. Translated by W. E. Weisflog. Deventer, The Netherlands: Kluwer, 1979.

Krishna, Vern. Fundamentals of Canadian Income Tax, vol. 1, Personal Tax, 2nd ed. Toronto: Thomson Reuters, 2019.

Krishna, Vern. Income Tax Law, 2nd ed. Toronto: Irwin Law, 2012.

Krishna, Vern. “Using Beneficial Ownership to Prevent Treaty Shopping” (2009), 56 Tax Notes Int’l 537.

Li, Jinyan. “Beneficial Ownership in Tax Treaties: Judicial Interpretation and the Case for Clarity”, in Philip Baker and Catherine Bobbett, eds., Tax Polymath: A life in international taxation. Amsterdam: IBFD, 2011, 187.

Li, Jinyan. “‘Economic Substance’: Drawing the Line Between Legitimate Tax Minimization and Abusive Tax Avoidance” (2006), 54 Can. Tax J. 23.

Li, Jinyan, Arthur Cockfield and J. Scott Wilkie. International Taxation in Canada: Principles and Practices, 2nd ed. Markham, Ont.: LexisNexis, 2011.

Li, Jinyan, and Arthur Cockfield, with J. Scott Wilkie. International Taxation in Canada: Principles and Practices, 4th ed. Toronto: LexisNexis, 2018.

Li, Jinyan, and Francesco Avella. “Article 13: Capital Gains”, in Global Tax Treaty Commentaries, last reviewed May 30, 2020 (online: https://research.ibfd.org/#/).

Loomer, Geoffrey T. “Tax Treaty Abuse: Is Canada Responding Effectively?”, in Oxford University Centre for Business Taxation, Working Paper No. 09/05, October 2008, revised March 2009.

Malherbe, Philippe. Elements of International Income Taxation. Brussels: Bruylant, 2015.

Organisation for Economic Co‑operation and Development. Action 6 Prevention of tax treaty abuse (online: https://www.oecd.org/tax/beps/beps-actions/action6/).

Organisation for Economic Co‑operation and Development. Background Brief: Inclusive Framework on BEPS, January 2017 (online: https://www.oecd.org/tax/beps/background-brief-inclusive-framework-on-beps.pdf).

Organisation for Economic Co‑operation and Development. Committee on Fiscal Affairs. Model Double Taxation Convention on Income and on Capital. Paris, 1977.

Organisation for Economic Co‑operation and Development. Committee on Fiscal Affairs. Model Tax Convention on Income and on Capital: Condensed Version. Paris, 1998.

Organisation for Economic Co‑operation and Development. Glossary of Tax Terms (online: https://www.oecd.org/ctp/glossaryoftaxterms.htm), “treaty shopping”.

Organisation for Economic Co‑operation and Development. Model Tax Convention on Income and on Capital: Condensed Version. Paris, 2003.

Organisation for Economic Co‑operation and Development. Model Tax Convention on Income and on Capital: Condensed Version. Paris, 2017.

Rosenbloom, H. David. “Tax Treaty Abuse: Policies and Issues” (1983), 15 Law & Pol’y Int’l Bus. 763.

Sasseville, Jacques. “Implementation of the General Anti-Avoidance Rule”, in Corporate Management Tax Conference, Income Tax Enforcement, Compliance, and Administration. Toronto: Canadian Tax Foundation, 1988, 4:1.

Simontacchi, Stefano. “Immovable Property Companies as Defined in Article 13(4) of the OECD Model” (2006), 60 Bull. Int’l Tax’n 29.

Simontacchi, Stefano. Taxation of Capital Gains under the OECD Model Convention: With special regard to Immovable Property, in Series on International Taxation, vol. 29. Alphen aan den Rijn, The Netherlands: Kluwer Law International, 2007.

United Nations. Model Double Taxation Convention between Developed and Developing Countries. U.N. Doc. ST/ESA/PAD/SER.E/21. New York, 2001.

United Nations. Committee of Experts on International Cooperation in Tax Matters. Report on the fourth session (20‑24 October 2008). U.N. Doc. E/2008/45 (Supp.). New York, 2008.

Vogel, Klaus. “Double Tax Treaties and Their Interpretation” (1986), 4 Int’l Tax & Bus. Law. 1.

Ward, David A. “Principles To Be Applied in Interpreting Tax Treaties” (1977), 25 Can. Tax J. 263.

Ward, David A., et al. The Interpretation of Income Tax Treaties with Particular Reference to the Commentaries on the OECD Model. Kingston: International Fiscal Association (Canadian Branch), 2005.

                    APPEAL from a judgment of the Federal Court of Appeal (Webb, Near and Locke JJ.A.), 2020 FCA 43, [2020] 5 C.T.C. 193, 2020 D.T.C. 5021, [2020] F.C.J. No. 204 (QL), 2020 CarswellNat 314 (WL Can.), affirming a decision of Hogan J., 2018 TCC 152, [2019] 5 C.T.C. 2183, 2018 D.T.C. 1120, [2018] T.C.J. No. 124 (QL), 2018 CarswellNat 4615 (WL Can.). Appeal dismissed, Wagner C.J. and Rowe and Martin JJ. dissenting.

                    Michael Taylor and Natalie Goulard, for the appellant.

                    Matthew G. Williams and E. Rebecca Potter, for the respondent.

 

The judgment of Abella, Moldaver, Karakatsanis, Côté, Brown and Kasirer JJ. was delivered by

 

                    Côté J. —

TABLE OF CONTENTS
Paragraph
I.      Overview	1
II.    Background	11
III.   Judicial History	18
A.    Tax Court of Canada, 2018 TCC 152, [2019] 5 C.T.C. 2183 (Hogan J.)	18
B.    Federal Court of Appeal, 2020 FCA 43, [2020] 5 C.T.C. 193 (Webb, Near and         Locke JJ.A.)	23
IV.   Issues	28
V.    Analysis	29
A.    General Anti-Avoidance Rule (“GAAR”)	29
B.    International Tax Treaties	34
(1)      General Principles	34
(2)      OECD Commentaries as Interpretative Aids	38
C.    Cautionary Preface to the GAAR Analysis	46
D.    First Step: Object, Spirit, and Purpose of the Relevant Provisions	50
(1)      Residence (Arts. 1 and 4(1))	52
(2)      Carve-Out from Source-Based Capital Gains Tax (“Business Property Exemption” (Art. 13(4) and (5))	68
E.     Second Step: Abusiveness of the Transaction	90
VI.   Conclusion	97

I.               Overview

[1]                              The principles of predictability, certainty, and fairness and respect for the right of taxpayers to legitimate tax minimization are the bedrock of tax law. In the context of international tax treaties, respect for negotiated bargains between contracting states is fundamental to ensure tax certainty and predictability and to uphold the principle of pacta sunt servanda, pursuant to which parties to a treaty must keep their sides of the bargain.

[2]                              Section 245 of the Income Tax Act, R.S.C. 1985, c. 1 (5th Supp.) (“Act”), known as the general anti-avoidance rule (“GAAR”), acts as a legislative limit on tax certainty by barring abusive tax avoidance transactions, including those in which taxpayers seek to obtain treaty benefits that were never intended by the contracting states. This intention is found by going behind the text of the provisions under which a tax benefit is claimed in order to determine their object, spirit, and purpose. In the bilateral treaty context, there are two sovereign states whose intentions are relevant; a robust analysis must take both into consideration in order to give proper effect to the tax treaty as a carefully negotiated instrument.

[3]                              In this case, the appellant, Her Majesty The Queen, as represented by the Minister of National Revenue (“Minister”), submits that the transaction at issue abused the Convention between the Government of Canada and the Government of the Grand Duchy of Luxembourg for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and on Capital, Can. T.S. 2000 No. 22 (“Treaty”). According to the Minister, the drafters did not intend the Treaty to benefit residents without “sufficient substantive economic connections” to their state of residence (A.F., at para. 100). In the view of the respondent, Alta Energy Luxembourg S.A.R.L. (“Alta Luxembourg”), the Minister has failed to discharge her burden of establishing that the object, spirit, or purpose of the provisions was frustrated or defeated.

[4]                              In my view, the Minister is asking this Court to use the GAAR to change the result, not by interpreting the provisions of the Treaty through a unified textual, contextual, and purposive analysis, but by fundamentally altering the criteria under which a person is entitled to the benefits of the Treaty, thus frustrating the certainty and predictability sought by the drafters.

[5]                              Tax treaties are replete with choices. One key choice made by Canada and Luxembourg was to deviate from the Organisation for Economic Co‑operation and Development (“OECD”) Model Tax Convention on Income and on Capital (“OECD Model Treaty”)^[1] by including a specific carve-out provision for immovable property, also called the business property exemption. This carve-out allocates to a person’s residence state the right to tax capital gains realized on the disposition of shares or other similar interests deriving their value principally from immovable property used in a corporation’s business. The rationale of the carve-out is not connected to the theory of economic allegiance. In fact, this provision is a clear departure from this theory, for the source state normally has the greater economic claim to tax income derived from immovable property or a business situated within its territory.

[6]                              Canada’s decision to forego its right to tax such capital gains realized in Canada was based on economic considerations broader than generating tax revenues. Tax law is designed not only to bring revenues into a state’s coffers but also to incentivize or disincentivize certain behaviours (Canada Trustco Mortgage Co. v. Canada, 2005 SCC 54, [2005] 2 S.C.R. 601, at para. 53). Indeed, in agreeing to include the carve-out in the Treaty, Canada sought to encourage investments by Luxembourg residents in business assets embodied in immovable property located in Canada (e.g., mines, hotels, or oil shales) and to reap the ensuing economic benefits. This incentive was never intended to be limited to Luxembourg residents with “sufficient substantive economic connections” to Luxembourg. Internationally, residency typically does not depend on the existence of such connections; formal criteria for residency are just as well accepted as factual criteria.

[7]                              In this case, Alta Luxembourg made exactly such an investment. It is a resident of Luxembourg and, as such, is exempt from Canadian taxes on the capital gain realized on the disposition of shares of its wholly owned Canadian subsidiary.

[8]                              In my respectful view, my colleagues Rowe and Martin JJ. undertake their analysis as though the Treaty were a simple statute rather than a freely negotiated bargain whose interpretation must reflect the intentions of the parties that drafted it. Canada understood that it was dealing with a low-tax jurisdiction, and, in recognition of this reality, it agreed to specific terms in the Treaty, such as the business property exemption. In this way, Canada effectively agreed to give up its right to tax certain entities incorporated in Luxembourg in exchange for the jobs and economic opportunities that the business property exemption would promote. This decision can hardly be questioned.

[9]                              In raising the GAAR, Canada is now seeking to revisit its bargain in order to secure both foreign investments and tax revenues. But if the GAAR is to remain a robust tool, it cannot be used to judicially amend or renegotiate a treaty.

[10]                          For the reasons that follow, I agree with the courts below that the Minister has not discharged her burden of proving abusive tax avoidance. Therefore, I would dismiss the appeal.

II.            Background

[11]                          In April 2011, Alta Resources LLC, a Texas-based oil and gas firm, and Blackstone Group LP (“Blackstone”), a New York-based private equity firm, founded Alta Energy Partners, LLC, a Delaware limited liability company, for the purpose of acquiring and developing unconventional oil and natural gas properties in North America. One such property was the Duvernay shale formation in northwestern Alberta. Alta Energy Partners Canada Ltd. (“Alta Canada”), a wholly owned Canadian subsidiary of the Delaware limited liability company, was incorporated in order to carry on that business. Alta Canada invested almost $300 million in its Canadian business through its acquisition of the oil and natural gas drill and recovery rights in certain lands in Alberta.

[12]                          A restructuring of Alta Canada was undertaken in 2012. As part of the restructuring, Alta Luxembourg was incorporated under the laws of Luxembourg to hold interests in Luxembourg and foreign companies. Prior to the restructuring, Blackstone’s counsel obtained a ruling from the Luxembourg tax authorities that the restructuring was in compliance with tax legislation and administrative policies in Luxembourg. The shares of Alta Luxembourg were issued to a new Canadian partnership formed in Alberta, Alta Energy Canada Partnership (“Partnership”). On the same day, the Delaware limited liability company sold all of its shares of Alta Canada to Alta Luxembourg. This was a taxable transaction in Canada under the Act, as more than 50 percent of the value of the shares was derived from Canadian resource properties.

[13]                          In August 2013, Alta Luxembourg agreed to sell its shares of Alta Canada to Chevron Canada Ltd. When the sale closed on September 10, 2013, for $679,712,251.45, Alta Luxembourg realized a capital gain in excess of $380 million on the disposition. Pursuant to a direction to pay, Alta Luxembourg directed its proceeds from the sale (less an amount paid to the Minister) to the Partnership. In exchange, the Partnership issued promissory notes to Alta Luxembourg, which were set off, in part, by an existing interest-free loan and profit‑participating loan. In other words, Alta Luxembourg did not receive any of the sale proceeds. Following the disposition of its shares of Alta Canada, Alta Luxembourg did not conduct any other business or hold any other investments.

[14]                          All tax was reported. Alta Luxembourg’s capital gain was reported to the Luxembourg tax authorities and was subject to full taxation by them under their domestic laws.

[15]                          In its Canadian tax return filed for the 2013 taxation year, Alta Luxembourg claimed an exemption from Canadian tax on the basis that the gain was not included in its “taxable income earned in Canada” under s. 115(1)(b) of the Act because the shares were “treaty-protected property” under art. 13(4) and (5) of the Treaty. These provisions create a carve‑out for residents of Luxembourg from Canadian tax arising from a capital gain on the alienation of “shares . . . the value of which . . . is derived principally from immovable property situated in [Canada]” and “in which the business of the company . . . was carried on” (art. 13(4) of the Treaty).

[16]                          The Minister denied the treaty exemption, and Alta Luxembourg appealed to the Tax Court of Canada.

[17]                          In their lengthy Statement of Agreed Facts, the parties make important concessions. First and foremost, the Minister agrees that Alta Luxembourg is a resident of Luxembourg for the purposes of the Treaty. The Minister and Alta Luxembourg also agree that Alta Canada was a “principal business corporation” pursuant to s. 66(15) of the Act, that its shares were taxable Canadian property within the meaning of s. 248(1) of the Act and that the series of restructuring transactions and the sale of the shares of Alta Canada to Chevron were an “avoidance transaction” as defined in s. 245(3) of the Act.

III.         Judicial History

A.           Tax Court of Canada, 2018 TCC 152, [2019] 5 C.T.C. 2183 (Hogan J.)

[18]                          Before the Tax Court, the Minister raised two arguments. First, the Minister argued that Alta Canada did not carry on business in the immovable property in question, such that the business property exemption in art. 13(4) of the Treaty did not apply. In the alternative, if the shares did qualify as treaty-protected property, the Minister argued that the GAAR should apply.

[19]                          Hogan J. found that Alta Canada carried on business in the immovable property; therefore, the carve-out in art. 13(4) applied, and Alta Canada’s shares were treaty-protected property for the purposes of the Act. With respect to the GAAR, the parties agreed that the restructuring was an avoidance transaction that resulted in a tax benefit. Thus, the Tax Court had to determine if the GAAR applied, that is, whether the avoidance transaction resulted in a misuse or abuse of the provisions of the Act or the Treaty.

[20]                          With respect to the abuse analysis under the GAAR, Hogan J. held that the avoidance transaction did not result in an abuse of ss. 2(3), 38, 39, 115(1)(b) and 248(1) of the Act, or of the Act as a whole. In his view, the provisions of the Act had operated in the manner intended by Parliament. Without citing this Court’s decision in Canada Trustco, Hogan J. followed the two-step approach it established, acknowledging that a tax treaty, as an international convention, should be given a liberal interpretation with a view to implementing the true intention of the parties.

[21]                          Hogan J. conducted a textual, contextual, and purposive analysis of arts. 1, 4 and 13 of the Treaty and held that their rationale is “to exempt residents of Luxembourg from Canadian taxation where there is an investment in immovable property used in a business” (para. 100). He relied in part on the 2017 OECD Model Treaty, on which the Treaty is modeled, and its Commentaries as interpretive aids. He observed that the business property exemption in art. 13(4) of the Treaty does not exist in the OECD Model Treaty, which demonstrates an intention to depart from the Model Treaty in order to attract foreign investment in business property situated in Canada. Despite the Minister’s objections, Hogan J. found nothing improper about Alta Luxembourg, a single‑purpose holding corporation resident in Luxembourg, availing itself of the benefits of the Treaty. In his view, the Minister was seeking to achieve through the courts the same result with the GAAR as was intended by the Minister of Finance’s proposed rule against tax treaty shopping, but the GAAR could not be used in this way.

[22]                          Hogan J. held that the overall result of the transaction was not contrary to the rationale of arts. 1, 4 and 13 because the “significant investments of [Alta Luxembourg] to de-risk the Duvernay shale constitute an investment in immovable property used in a business” (para. 100). Therefore, the GAAR did not preclude Alta Luxembourg from claiming the exemption provided for in art. 13(5) of the Treaty, and the matter was referred back to the Minister for reconsideration and reassessment.

B.            Federal Court of Appeal, 2020 FCA 43, [2020] 5 C.T.C. 193 (Webb, Near and Locke JJ.A.)

[23]                          The Minister did not appeal the Tax Court’s finding that Alta Luxembourg satisfied the requirements for the business property exemption. Thus, the sole issue before the Federal Court of Appeal was whether the GAAR applied because of an abuse of the Act or the Treaty.

[24]                          Webb J.A., writing for a unanimous court, dismissed the appeal. In his analysis of the object, spirit, and purpose of the relevant provisions, Webb J.A. noted that most of the Minister’s submissions were in reference to general principles and failed to identify any clear rationale for the provisions at issue: arts. 1, 4 and 13(4) of the Treaty. Relying on R. v. MIL (Investments) S.A., 2007 FCA 236, [2007] 4 C.T.C. 235, Webb J.A. concluded “that the object, spirit and purpose of the relevant provisions of the [Treaty] is reflected in the words as chosen by Canada and Luxembourg. Since the provisions operated as they were intended to operate, there was no abuse” (para. 80). He also determined that Commentaries on the OECD Model Treaty published subsequently to the signature and ratification of the Treaty were “of little assistance in determining the rationale for the exemption” (para. 36).

[25]                          In addition to finding that the Minister had failed to identify any clear rationale for arts. 1, 4 and 13(4) of the Treaty, Webb J.A. found that the Minister’s five further submissions did not withstand scrutiny, as they all added qualifications to or modified the words of the Treaty. The Minister’s submissions changed the identity of who would qualify for the exemption from “residents” to “investors” and added the qualification that an “entity” had to have the potential to earn income in Luxembourg in order to be considered a resident of Luxembourg. In Webb J.A.’s view, the fact that the net effect of the profit‑participating loan was that Alta Luxembourg never realized taxable income in Luxembourg “is a matter for the Luxembourg tax authorities” (para. 57). Additionally, there was no underlying requirement that the exemption benefit only persons with commercial or economic ties to Luxembourg, nor was the residence of the partners of Alta Luxembourg’s sole shareholder relevant to the analysis. He found that these qualifications were not included in the exemption in the Treaty, even though they could easily have been added.

[26]                          Webb J.A. declined to find that treaty shopping is abusive, agreeing with the Tax Court judge in MIL (Investments) S.A. v. R., 2006 TCC 460, [2006] 5 C.T.C. 2552 (“MIL (TCC)”), at para. 69, that “[t]here is nothing inherently proper or improper with selecting one foreign regime over another” and that, though “the selection of a low tax jurisdiction may speak persuasively as evidence of a tax purpose for an alleged avoidance transaction, . . . the shopping or selection of a treaty to minimize tax on its own cannot be viewed as being abusive”.

[27]                          In short, Webb J.A. found that “the object, spirit and purpose of the relevant provisions of the [Treaty] is reflected in the words as chosen by Canada and Luxembourg. Since the provisions [had] operated as they were intended to operate, there was no abuse” (para. 80). The appeal was therefore dismissed.

IV.         Issues

[28]                          The Minister has accepted that Alta Luxembourg is a resident of Luxembourg for the purposes of the Treaty and that a business was being carried on in the immovable property in question. Furthermore, Alta Luxembourg has admitted the existence of a tax benefit and an avoidance transaction. Therefore, the only element in dispute is the abusive nature of the transaction, which raises the following issues:

(a)   What are the object, spirit, and purpose of the relevant provisions of the Treaty?

 

(b)   Did the courts below err in concluding that the avoidance transaction in this case did not result in an abuse of those provisions?

 

V.           Analysis

A.           General Anti-Avoidance Rule (“GAAR”)

[29]                          Like all statutes, tax legislation must be interpreted by conducting a “textual, contextual and purposive analysis to find a meaning that is harmonious with the Act as a whole” (Canada Trustco, at para. 10). However, where tax provisions are drafted with “particularity and detail”, a largely textual interpretation is appropriate in light of the well-accepted Duke of Westminster principle that “taxpayers are entitled to arrange their affairs to minimize the amount of tax payable” (Canada Trustco, at para. 11, citing Commissioners of Inland Revenue v. Duke of Westminster, [1936] A.C. 1 (H.L.)). This principle, derived from the rule of law, has been deemed the “foundation stone of Canadian law on tax avoidance” (B. J. Arnold, “Reflections on the Relationship Between Statutory Interpretation and Tax Avoidance” (2001), 49 Can. Tax J. 1, at p. 3).

[30]                          This established principle was affected by the enactment of s. 245 of the Act, also known as the GAAR, which “superimposed a prohibition on abusive tax avoidance, with the effect that the literal application of provisions of the Act may be seen as abusive in light of their context and purpose” (Canada Trustco, at para. 1). Thus, if the Minister can establish abusive tax avoidance under the GAAR, s. 245 of the Act will apply to deny the tax benefit even where the tax arrangements are consistent with a literal interpretation of the relevant provisions (Copthorne Holdings Ltd. v. Canada, 2011 SCC 63, [2011] 3 S.C.R. 721, at para. 66). The GAAR applies both to the abuse of provisions found in the Act and to the abuse of provisions found in a tax treaty (s. 245(4)(a)(i) and (iv) of the Act; s. 4.1 of the Income Tax Conventions Interpretation Act, R.S.C. 1985, c. I-4).

[31]                          Applying the GAAR involves a three-part process meant to determine: (1) whether there is a tax benefit arising from a transaction; (2) whether the transaction is an avoidance transaction; and (3) whether the avoidance transaction is abusive (Canada Trustco, at para. 17). As mentioned above, the third part is the only one in issue before this Court. To determine whether a transaction is abusive, this Court has set out a two-step inquiry (Canada Trustco, at paras. 44 and 55). Under the first step, the provisions relied on for the tax benefit are interpreted to determine their object, spirit, and purpose. The second step is to undertake a factual analysis to determine whether the avoidance transaction at issue is consistent with or frustrates the object, spirit, and purpose of the provisions.

[32]                          The onus rests on the Minister to demonstrate the object, spirit, and purpose of the relevant provisions and to establish that allowing Alta Luxembourg the benefit of the exemption would be a misuse or an abuse of the provisions (Canada Trustco, at para. 65). Abusive tax avoidance occurs “when a taxpayer relies on specific provisions of the Income Tax Act in order to achieve an outcome that those provisions seek to prevent” or when a transaction “defeats the underlying rationale of the provisions that are relied upon” (Canada Trustco, at paras. 45; see also para. 57; Lipson v. Canada, 2009 SCC 1, [2009] 1 S.C.R. 3, at para. 40). Abusive tax avoidance can also occur when an arrangement “circumvents the application of certain provisions, such as specific anti‑avoidance rules, in a manner that frustrates or defeats the object, spirit or purpose of those provisions” (para. 45).

[33]                          Canada Trustco recognized that the line between legitimate tax minimization and abusive tax avoidance is “far from bright” (para. 16). As a result, “[i]f the existence of abusive tax avoidance is unclear, the benefit of the doubt goes to the taxpayer” (Canada Trustco, at para. 66; see also Copthorne, at para. 72).

B.            International Tax Treaties

(1)          General Principles

[34]                          In R. v. Melford Developments Inc., [1982] 2 S.C.R. 504, at p. 513, this Court applied the principle that tax treaties do not themselves levy new taxes, they simply authorize the contracting parties to do so. Reciprocity is a fundamental principle underlying tax treaties, as they confer rights and impose obligations on each of the contracting states. Hogan J. observed that “[p]arties to a tax treaty are presumed to know the other country’s tax system when they negotiate a tax treaty; they are presumed to know the tax consequences of a tax treaty when they negotiate amendments to that treaty” (para. 84). This only makes sense.

[35]                          The objective of tax treaties, broadly stated, is to govern the interactions between national tax laws in order to facilitate cross-border trade and investment. One of the most important operational goals is the elimination of double taxation, where the same source of income is taxed by two or more states without any relief. If left unchecked, double taxation risks creating barriers to international trade and investment, which are vital in a globalized economy. Thus, many substantive provisions of the OECD Model Treaty, a model for numerous bilateral tax treaties, are directed to achieving this goal and resolving conflicting claims between residence-based taxation and source-based taxation.

[36]                          Another important consideration is the dual nature — contractual and statutory — of tax treaties. Consideration of the contractual element is crucial to the application of the GAAR because it focuses the analysis on whether the particular tax planning strategy is consistent with the compromises reached by the contracting states. As noted by international tax law scholars Jinyan Li and Arthur Cockfield:

Whether the particular outcome of tax planning is defensible may depend on the understanding of the “bargain” struck by the two treaty partner countries. Every dispute involving the application of a tax treaty needs to ask the question of whether and how one treaty partner can dispute or should be allowed to upset the “bargain” struck in its own national interest that inheres in the treaty “contract”. Despite the offence that one treaty partner may take, in retrospect, to how a treaty provision is applied, the question remains: Might the particular outcome be one that the other treaty partner foresaw or reflect the “contractual intention” of the other treaty partner? After all, the “bargain” was entered into by the parties out of mutual self-interest. This is particularly relevant in applying general anti‑avoidance rules. [Emphasis added.]

 

(J. Li and A. Cockfield, with J. S. Wilkie, International Taxation in Canada: Principles and Practices (4th ed. 2018), at p. 376)

[37]                          As tax treaties are treaties, their interpretation is governed by the Vienna Convention on the Law of Treaties, Can. T.S. 1980 No. 37 (“Vienna Convention”), but the methodology prescribed is not radically different from the modern principle applicable to domestic statutes in Canada — that is, one must consider the ordinary meaning of the text in its context and in light of its purpose (art. 31(1) of the Vienna Convention; Crown Forest Industries Ltd. v. Canada, [1995] 2 S.C.R. 802, at para. 43; Stubart Investments Ltd. v. The Queen, [1984] 1 S.C.R. 536, at p. 578). However, unlike statutes, treaties must be interpreted “with a view to implementing the true intentions of the parties” (J. N. Gladden Estate v. The Queen, [1985] 1 C.T.C. 163 (F.C.T.D.), at p. 166, quoted approvingly in Crown Forest, at para. 43). The national self-interest of each contracting state must be reconciled in the interpretive process in order to give full effect to the bargain codified by the treaty. This principle applies with equal force where a court is engaged in the process of ascertaining a treaty’s “object, spirit, and purpose” as part of the GAAR framework.

(2)          OECD Commentaries as Interpretative Aids

[38]                          Article 31 of the Vienna Convention permits courts to consider contextual factors such as other agreements and instruments made by parties in connection with a treaty. In my view, the OECD Model Treaty and its Commentaries are relevant to the interpretation of treaties based on that model. The introduction to the OECD Model Treaty indicates that the Commentaries “can . . . be of great assistance in the application and interpretation of the conventions and, in particular, in the settlement of any disputes”, and this Court has affirmed the “high persuasive value” of the OECD Model Treaty and its Commentaries (“Introduction” to the OECD Model Treaty (1998, 2003 and 2017), at para. 29; Crown Forest, at para. 55; see also D. A. Ward, “Principles To Be Applied in Interpreting Tax Treaties” (1977), 25 Can. Tax J. 263, at p. 268). However, the relevance of Commentaries released subsequent to the signing of a treaty is disputed (see, e.g., “Introduction” to the OECD Model Treaty (1998, 2003 and 2017), at para. 35; MIL (TCC), at para. 83; Cudd Pressure Control Inc. v. R., [1999] 1 C.T.C. 1 (F.C.A.), at para. 28, per McDonald J.A.; SA Andritz, No. 233894, Conseil d’État (Section du Contentieux), December 30, 2003 (France); Li and Cockfield, at p. 57).

[39]                          In the instant case, the Minister relies on revisions to the Commentaries on the OECD Model Treaty that were published in 2003 and 2017, several years after Canada and Luxembourg negotiated the Treaty. In the 2003 Commentaries, treaty shopping is characterized as an abuse of the concept of residence, whereas previous Commentaries published at the time the Treaty was signed were silent on this question. A revision to the 2017 Commentaries, made in connection with the addition of a new art. 29 to the OECD Model Treaty, provides that legal residency alone is not an automatic entitlement to all benefits under a tax treaty.

[40]                          While revisions to the Commentaries are relevant to tax treaty interpretation, the key issue is the weight that they should receive. Although some scholars submit that the OECD has a tendency of revising the Commentaries too often and too dramatically, thereby sometimes diverging from the original intentions of the parties, I am not prepared to reject all subsequent Commentaries as interpretative aids (see Li and Cockfield, at p. 57; P. Malherbe, Elements of International Income Taxation (2015), at pp. 49-50). I instead prefer the nuanced approach adopted by the Federal Court of Appeal in Prévost Car Inc. v. Canada, 2009 FCA 57, [2010] 2 F.C.R. 65.

[41]                          Indeed, in Prévost Car, the Federal Court of Appeal held that subsequent Commentaries expanding or clarifying notions already captured by the OECD Model Treaty are relevant, but not those that extend the scope of provisions in a manner that could not have been considered by the drafters (paras. 10-12; see also Li and Cockfield, at p. 57). Thus, while later amendments to the Commentaries are not part of the context as defined in art. 31(2) of the Vienna Convention, given that such amendments were not made “in connexion with the conclusion of the treaty”, they may play a role under art. 31(3), which refers to “[a]ny subsequent agreement between the parties regarding the interpretation of the treaty or the application of its provisions” and “[a]ny subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation”.

[42]                          In this case, I am of the view that the 2003 and 2017 Commentaries do not reflect the intentions of the drafters of the Treaty. The extensive revisions made to the Commentaries in 2003 purported to clarify the relationship between tax treaties and domestic anti-avoidance rules, and, in particular, one of the revisions was made to include the prevention of tax avoidance as a purpose of such treaties. The changes were not mere clarifications and have been described as being “created out of thin air by the OECD in 2003” and as “a significant change in the stated attitude of the OECD to the relationship between tax treaties and tax avoidance” (B. J. Arnold, “Tax Treaties and Tax Avoidance: The 2003 Revisions to the Commentary to the OECD Model” (2004), 58 I.B.F.D. Bulletin 244, at pp. 249 and 260).

[43]                          Using the Federal Court of Appeal’s language in Prévost Car (at para. 12), the 2003 Commentaries do not elicit, but rather contradict, the views previously expressed. When Canada and Luxembourg signed the Treaty in 1999, the applicable Commentaries indicated that anti-abuse measures, to be effective, had to be included in a treaty (“Commentary on Article 1” of the 1998 OECD Model Treaty, at para. 21). Further, they referred to the principle of pacta sunt servanda, which supports the position that where nothing in a treaty speaks directly to fiscal avoidance, there is a strong argument that the treaty partners negotiated the treaty not intending such rules to apply (“Commentary on Article 1” of the 1998 OECD Model Treaty, at paras. 11‑26; D. A. Ward et al., The Interpretation of Income Tax Treaties with Particular Reference to the Commentaries on the OECD Model (2005), at pp. 91‑92).

[44]                          Moreover, interpreting art. 1 of the Treaty with reference to the 2003 Commentaries would overlook Luxembourg’s registered observation on the “Commentary on Article 1” of the 2003 OECD Model Treaty. That observation reads as follows:

Luxembourg does not share the interpretation in paragraphs 9.2, 22.1 and 23 which provide that there is generally no conflict between anti-abuse provisions of the domestic law of a Contracting State and the provisions of its tax conventions. Absent an express provision in the Convention, Luxembourg therefore believes that a State can only apply its domestic anti-abuse provisions in specific cases after recourse to the mutual agreement procedure. [para. 27.6]

In effect, even if the Minister were able to rely on the Commentaries postdating the Treaty, they would be of no assistance because Luxembourg’s observation expresses disagreement with the “Commentary on Article 1” of the OECD Model Treaty, which includes the anti-abuse commentary (Ward et al., at p. 64; “Commentary on Article 1” of the 2003 OECD Model Treaty, at p. 7).

[45]                          It follows, then, that in the interpretation of art. 1 of the Treaty, Commentaries on art. 1 of the OECD Model Treaty that postdate the Treaty cannot be relied on to introduce terms that modify the Treaty. Not only would this effectively amend the Treaty in a manner not agreed upon by the parties, but it would also usurp the role of the Governor in Council by allowing for judicial amendment of bilateral treaties against the expressed wishes of the contracting states.

C.            Cautionary Preface to the GAAR Analysis

[46]                          Before I proceed, it is important to sound some notes of caution.

[47]                          First and foremost, tax avoidance is not tax evasion, and there is no suggestion by either party that the transaction in this case was evasive. In addition, tax avoidance should not be conflated with abuse. Even if a transaction was designed for a tax avoidance purpose and not for a bona fide non-tax purpose, such as an economic or commercial purpose, it does not mean that it is necessarily abusive within the meaning of the GAAR (Canada Trustco, at paras. 36 and 57; see also Lipson, at para. 38). The purpose of a transaction is relevant mainly to characterize it as either an avoidance transaction or a bona fide transaction and, specifically, to assess the abusive nature of the transaction. In their factual analysis, courts may consider whether an avoidance transaction was “motivated by any economic, commercial, family or other non-tax purpose” (Canada Trustco, at para. 58). However, a finding that a bona fide non-tax purpose is lacking, taken alone, should not be considered conclusive evidence of abusive tax avoidance. Justices Rowe and Martin are taking exactly that approach, and it colours their entire analysis. Moreover, such a finding should not be allowed to impair the proper interpretation of the relevant provisions in a manner that makes substantive economic connections or the presence of a bona fide non-tax purpose a condition precedent to every tax benefit; the goal is to ensure the relevant provisions are properly interpreted in light of their context and purpose (Canada Trustco, at para. 62).

[48]                          Second, it is also important to distinguish what is immoral from what is abusive. It is true, as reiterated in Copthorne, that the GAAR is a legislative measure by which “Parliament has conferred on the court the unusual duty of going behind the words of the legislation to determine the object, spirit or purpose of the provision or provisions relied upon by the taxpayer” (para. 66). But, in Copthorne, Rothstein J. was quick to note the limits to that legislative mandate. In contrast to what my colleagues are proposing, Rothstein J. observed that courts should not infuse the abuse analysis with “a value judgment of what is right or wrong nor with theories about what tax law ought to be or ought to do” (para. 70). Taxpayers are allowed to minimize their tax liability to the full extent of the law and to engage in “creative” tax avoidance planning, insofar as it is not abusive within the meaning of the GAAR (para. 65). Therefore, even though one may consider treaty shopping in tax havens to be immoral, this is not determinative of a finding of abuse.

[49]                          Finally, the abuse analysis is not meant to be a “search for an overriding policy of the Act that is not based on a unified, textual, contextual and purposive interpretation of the specific provisions in issue” (Canada Trustco, at para. 41). The focus of the interpretation is on the object, spirit, and purpose of the specific provisions and not on the broader policy objective of the Act or of a particular tax treaty. Therefore, policy objectives such as “avoiding double taxation” and “encouraging trade and investment” that are found in bilateral tax treaties cannot be invoked to override the wording of the provisions in issue.

D.           First Step: Object, Spirit, and Purpose of the Relevant Provisions

[50]                          As mentioned above, the first step of the abuse analysis is to ascertain the object, spirit, and purpose of the relevant provisions. Because this is a question of treaty interpretation, however, this must be done with a view to implementing the true intentions of the parties. This is a question of law and the analysis of this first step is therefore subject to the correctness standard (Canada Trustco, at para. 44).

[51]                          The Minister’s submissions centre on an alleged abuse by Alta Luxembourg of arts. 1, 4(1) and 13(4) and (5) of the Treaty. I analyze these provisions in two separate groups for the purposes of the first step: first, arts. 1 and 4(1), which pertain to resident status; and second, art. 13(4) and (5), under which the right to tax the capital gain at issue is allocated to the residence state.

(1)          Residence (Arts. 1 and 4(1))

[52]                          Residence is at the core of bilateral tax treaties, given that access to treaty benefits is normally reserved to persons residing in one or both of the contracting states. The text of arts. 1 and 4(1) of the Treaty also makes residence central to the application of the Treaty. Indeed, the residency requirement established in art. 1 of the Treaty is modeled on the OECD Model Treaty:

This Convention shall apply to persons who are residents of one or both of the Contracting States.

[53]                          Article 4(1) elaborates on the definition of “residence” under the Treaty:

For the purposes of this Convention, the term “resident of a Contracting State” means any person who, under the laws of that State, is liable to tax therein by reason of that person’s domicile, residence, place of management or any other criterion of a similar nature. This term also includes a Contracting State or a political subdivision or local authority thereof or any agency or instrumentality of any such State, subdivision or authority. This term, however, does not include any person who is liable to tax in that State in respect only of income from sources in that State.

According to this provision, a resident under the Treaty is a person who is liable to tax in one or both of the contracting states (Canada and Luxembourg) by reason of one of the connecting factors listed (i.e. domicile, residence, place of management or another similar criterion) (see Crown Forest, at paras. 23-25). I also note that the use of the word “means” in this provision indicates that the definition should be “construed as comprehending that which is specifically described or defined” and thus as setting out all requirements that must be met to be considered a resident under the Treaty (R. v. Hauser, [1979] 1 S.C.R. 984, at p. 1009, per Dickson J.; see also R. v. McLeod (1950), 97 C.C.C. 366 (B.C.C.A.), at pp. 371-72, quoting Dilworth v. Commissioner of Stamps, [1899] A.C. 99 (P.C.), at pp. 105‑6).

[54]                          In the context of corporations, the “liable to tax” requirement is met under the Treaty where the domestic law of a contracting state exposes the corporation to full tax liability on its worldwide income because it has its residence in that state (see Crown Forest, at paras. 40 and 45). Liability to full taxation is established by the nexus between that State and the corporation’s resident status. The “liable to tax” requirement is often described in terms that may perhaps appear misleading, such as “comprehensive taxation” or “full liability to tax”. These terms convey the idea that residents enjoying tax holidays may be more suspicious than others. In reality, this requirement is not concerned with whether the person claiming benefits is in fact subject to taxation. Being liable to tax is better understood as being “liable to be liable to tax”, meaning that taxes are a possibility, regardless of whether the person actually pays any (R. Couzin, Corporate Residence and International Taxation (2002), at p. 107; see also pp. 106 and 111). Therefore, corporate residents enjoying certain tax holidays, for example on capital gains, do not automatically lose their resident status under the Treaty because they are not subject to every possible form of taxation (Couzin, at pp. 110-11 and 150). This can be contrasted with fiscally transparent vehicles like partnerships that are not exempted from taxation but, rather, are not exposed to tax at all, as their income is taxed in the partners’ hands instead.

[55]                          Aside from the “liable to tax” requirement, the purpose of art. 4(1) is not to establish specific standards for defining residence. This provision expressly states that residence is to be defined by the laws of the contracting state of which the person claims to be a resident. This provision of the Treaty is modeled almost word for word on art. 4(1) of the 1998 OECD Model Treaty, whose Commentary also made it clear that the intention was to leave the core definition of residence to domestic law, not to bilateral tax treaties:

Conventions for the avoidance of double taxation do not normally concern themselves with the domestic laws of the Contracting States laying down the conditions under which a person is to be treated fiscally as “resident” and, consequently, is fully liable to tax in that State. They do not lay down standards which the provisions of the domestic laws on “residence” have to fulfil in order that claims for full tax liability can be accepted between the Contracting States. In this respect the States take their stand entirely on the domestic laws. [Emphasis added.]

 

(“Commentary on Article 4” of the 1998 OECD Model Treaty, at para. 4)

[56]                         Consideration of the context of the Treaty confirms this intention expressed in the Commentary. Indeed, this preference for leaving the meaning of residence to domestic law is totally consistent with the scheme of the Treaty. Most terms found in the Treaty are defined under domestic law and not by the Treaty itself. As emphasized by Professor Arnold, “[b]ecause the language of tax treaties is broad and general, it seems inevitable that recourse must be had to the domestic laws of the contracting states in order to provide flesh for the bare bones of the treaty” (B. J. Arnold, Reforming Canada’s International Tax System: Toward Coherence and Simplicity (2009), at p. 325). Although the Treaty does define residence in art. 4 and some other terms in arts. 3(1), 5, and 6(2), these definitions are far from exhaustive. In fact, the list of defined terms is rather scant, and important concepts, such as “business” and “profits”, take their meaning directly from domestic law. The importance of domestic law as a source of substantive content for the application of the Treaty is expressly spelled out in art. 3(2):

As regards the application of the Convention at any time by a Contracting State, any term not defined therein shall, unless the context otherwise requires, have the meaning that it has at that time under the law of that State for the purposes of the taxes to which the Convention applies, any meaning under the applicable tax laws of that State prevailing over a meaning given to the term under other laws of that State.

[57]                          Despite the clear pronouncement made in the Commentary above and the well-established preference for leaving important definitions to domestic law, and despite her admission that Alta Luxembourg is a resident of Luxembourg, the Minister argues that meeting the definition of resident under domestic law is not sufficient to qualify as a resident under the Treaty. According to the Minister, the benefits of the Treaty “are intended to be available only to persons who have sufficient substantive economic connections” to their state of residence (A.F., at para. 100 (emphasis added)). Mere formalistic or legal attachment to their state of residence would thus be insufficient.

[58]                          It is worth noting that the words “sufficient substantive economic connections” are conspicuous by their absence in the text of both arts. 1 and 4. Although the GAAR invites courts to go beyond the text to understand the object, spirit, and purpose of the provisions, there are limits to this exercise, especially when attempting to discern the intent of bilateral treaty partners. In the face of a complete absence of express words, the inclusion of an unexpressed condition must be approached with circumspection. It must be remembered that the text also plays an important role in ascertaining the purpose of a provision. The proper approach is one that unifies the text, context, and purpose, not a purposive one in search of a vague policy objective disconnected from the text (Canada Trustco, at para. 41).

[59]                          Nonetheless, I acknowledge that treaty partners do not have the unfettered liberty to alter or redefine residence as they wish for the purposes of a tax treaty. The broader context of international tax law and the law of treaties helps to understand what was within the contemplation of Canada and Luxembourg when they drafted arts. 1 and 4(1) of the Treaty. Pursuant to the principle of pacta sunt servanda, parties to a treaty must keep their sides of the bargain and perform their obligations in good faith (art. 26 of the Vienna Convention). Domestic law definitions of residence should therefore broadly correspond to international norms and not have the effect of redefining residence in a way “that takes the words unmistakably past their accepted usage” (Couzin, at p. 136), including the definitions of residence that were in effect in the two states at the time the Treaty was drafted.

[60]                          I pause here to observe that the definition of residence in Luxembourg law is consistent with international practice. Broadly speaking, there are two internationally recognized methods used to determine corporate residency: (1) the “place of incorporation” or “legal seat” rule, pursuant to which residence is determined by a purely formal criterion, that is, where the corporation was incorporated or has its legal seat; and (2) the “real seat” rule, pursuant to which residence depends on a combination of factual factors aimed at identifying the corporation’s place of effective management (R. S. Avi-Yonah, N. Sartori and O. Marian, Global Perspectives on Income Taxation Law (2011), at p. 130, quoting M. A. Kane and E. B. Rock, “Corporate Taxation and International Charter Competition” (2008), 106 Mich. L. Rev. 1229, at p. 1235). Luxembourg law grants resident status to corporations having either their legal seat or their central management in Luxembourg — two criteria consistent with these methods (Statement of Agreed Facts, A.R., vol. II, at p. 28, para. 122; Opinion on the Luxembourg tax residence of the company, A.R., vol. VII, at pp. 5 and 7, paras. 7.2 and 8.2.1). In the instant case, the parties agree that Alta Luxembourg is a resident of Luxembourg as its legal seat is located there (Statement of Agreed Facts, A.R., vol. II, at p. 28, para. 122).

[61]                          Interestingly, the “sufficient substantive economic connections” rationale put forward by the Minister bears similarities to the “real seat” rule emphasizing substance over form and seemingly rejects a formal, legalistic rule like the “place of incorporation” or “legal seat” rule. Thus, I understand the Minister’s submissions as suggesting that establishing residence merely on the basis of a formal criterion is insufficient to conform to the spirit of the rules of residence under the Treaty. Something more would be needed: some real connections to the country of residence. What the Minister’s submissions overlook, however, is that many of the world’s most developed economies — including Canada itself — accept and apply the “place of incorporation” or “legal seat” rule (Avi-Yonah, Sartori and Marian, pp. 130 and 133‑34; see s. 250(4)(a) of the Act). Although a formal criterion may sometimes be unable to capture the real location of a corporation’s economic activities, it nevertheless became widespread internationally because of its certainty and simplicity, considerations that are vital to a well-functioning tax system based on the rule of law and the Duke of Westminster principle (Li and Cockfield, at p. 77). Hence, the definition in Luxembourg law does not depart from accepted usage such that the bargain struck in the Treaty could be upheld only if Luxembourg residents claiming benefits have “sufficient substantive economic connections” to their country of residence.

[62]                          Given this broad international acceptance of formal residency, if the drafters had truly intended to include only corporations with “sufficient substantive economic connections” to their country of residence within the scope of the Treaty, they would have clearly signalled their intention to depart from a well-established criterion like the “place of incorporation” or “legal seat” rule. They would not have simply incorporated arts. 1 and 4(1) of the OECD Model Treaty, which reflect an international consensus, with no alteration. This indicates, in my view, that the object of arts. 1 and 4(1) is not to exclude all corporations with minimal economic connections to their country of residence, such as those whose residence is established solely on the basis of a formal, legal attachment. Access to the benefits of the Treaty by virtue of a domestic law definition of residence like the “legal seat” rule is therefore entirely consistent with the spirit of these provisions.

[63]                          Another contextual element further reinforces my opinion that the purpose of arts. 1 and 4 is not to reserve benefits to corporations with “sufficient substantive economic connections” to their country of residence: the inclusion of art. 28(3) in the Treaty, which denies benefits to certain Luxembourg holding companies. The “Commentary on Article 1” of the 1998 OECD Model Treaty specifically warned against the use of conduit companies, that is, “legal entit[ies] created in a State essentially to obtain treaty benefits that would not be available directly” (para. 9). Several solutions were proposed in the Commentary, but “no definitive texts [were] drafted” and “no strict recommendations [were] made” (para. 12). OECD members were left to choose which solution suited them.

[64]                          Chief among these solutions were the “look-through” approach and the exclusion approach. A “look-through” provision disallows treaty benefits to corporations that are resident in a contracting state but owned by residents of a third country (para. 13). When combined with a provision safeguarding bona fide business activities, a provision of this type has the effect of allowing treaty benefits to corporations owned by residents of a third country that conduct bona fide business activities in the contracting state of which they claim to be residents, while excluding those conducting “little substantive business activities” there (para. 14). The exclusion approach denies treaty benefits to certain types of companies enjoying special tax privileges that constitute harmful tax competition (para. 15). The main advantage of the exclusion approach over the “look-through” one is its clarity and simplicity (para. 16). It does not require an assessment of the substantive nature of business activities to determine whether they are bona fide or not, unlike the “look-through” approach or the approach focusing on the existence of “sufficient substantive economic connections” invoked by the Minister.

[65]                          At the time of the Treaty, Luxembourg was well known as an international tax haven exempting certain holding companies from most taxes, provided they had no direct presence in the local economy (see Grundy’s Tax Havens: A World Survey (4th ed. 1983), at pp. 2 and 157-64; W. H. Diamond et al., Tax Havens of the World (loose-leaf), vol. 2, ch. Luxembourg, at pp. 5-6). To deny the benefits of the Treaty to such holding companies with minimal economic connections to Luxembourg, Canada and Luxembourg preferred the exclusion approach over the “look-through” approach. Indeed, the two contracting states did not insert a “look-through” provision combined with a safeguarding provision, as Canada did in the treaties it entered into during the same period with Kazakhstan (1996) and Peru (2001) (art. 28(3) of the Convention between the Government of Canada and the Government of the Republic of Kazakhstan for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and on Capital, Can T.S. 1998 No. 13; art. 28(3) of the Convention between the Government of Canada and the Government of the Republic of Peru for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and on Capital, Can. T.S. 2002 No. 23). Canada and Luxembourg included instead a provision denying the benefits of the Treaty to certain holding companies established in Luxembourg, that is, art. 28(3):

The Convention shall not apply to holding companies within the meaning of the special Luxembourg laws (currently the Act of July 31, 1929 and the Grand Duchy Order of December 17, 1938) or any other similar law enacted in Luxembourg after the signature of the Convention, nor to companies subjected to similar fiscal laws in Luxembourg.

[66]                          Pursuant to the principle of implied exclusion, this choice made by the parties in favour of the exclusion approach — one that favours form over substance — should be understood as a rejection of the relevance of economic ties for delineating which corporations should be entitled to benefits and which should not. This leads me to conclude that the drafters intended to exclude a corporation with minimal economic connections to one of the contracting states only where the corporation is a holding company benefiting from Luxembourg’s international tax haven regime. In light of this clear intention to reject only Luxembourg holding companies and not every company with limited economic ties to its country of residence, I am even more persuaded that the spirit of arts. 1 and 4(1) was not to limit access to the benefits of the Treaty to corporations with “sufficient substantive economic connections” to their country of residence.

[67]                          In sum, the object, spirit, and purpose of arts. 1 and 4(1) are to allow all persons who are residents under the laws of one or both of the contracting states to claim benefits under the Treaty so long as their resident status could expose them to full tax liability (regardless of whether there is actual taxation). They are broadly consistent with international norms. This is normally the case for corporations that are residents by virtue of the “place of incorporation” or “legal seat” rule, unless they fall within the exclusion provided for in art. 28(3). As a result, I conclude that the spirit of these provisions is not to reserve the benefits of the Treaty to residents that have “sufficient substantive economic connections” to their country of residence.

(2)          Carve-Out from Source-Based Capital Gains Tax (“Business Property Exemption”) (Art. 13(4) and (5))

[68]                          The carve-out from source-based capital gains tax, also called the business property exemption, is provided for in art. 13(4) and (5). These provisions create an exception to source-based taxation for capital gains derived from the alienation of shares deriving their value principally from immovable property used in a corporation’s business and allocate to the residence state the exclusive right to tax such capital gains:

ARTICLE 13

 

Capital Gains

 

. . .

 

4.  Gains derived by a resident of a Contracting State from the alienation of:

 

(a)   shares (other than shares listed on an approved stock exchange in the other Contracting State) forming part of a substantial interest in the capital stock of a company the value of which shares is derived principally from immovable property situated in that other State; or

 

(b)   an interest in a partnership, trust or estate, the value of which is derived principally from immovable property situated in that other State,

 

may be taxed in that other State. For the purposes of this paragraph, the term “immovable property” does not include property (other than rental property) in which the business of the company, partnership, trust or estate was carried on; and a substantial interest exists when the resident and persons related thereto own 10 per cent or more of the shares of any class or the capital stock of a company.

 

5.  Gains from the alienation of any property, other than that referred to in paragraphs 1 to 4 shall be taxable only in the Contracting State of which the alienator is a resident.

[69]                          With respect to the object, spirit, and purpose of art. 13(4) and (5), the crux of the Minister’s argument turns on the theory of economic allegiance. In accordance with this theory, the Minister submits that the provisions of the Treaty distributing taxing rights seek to avoid double taxation by allocating these rights to the contracting state to which the income and the taxpayer are more closely connected.

[70]                          According to the Minister, the carve-out provided for in art. 13(4) exempts the capital gains at issue from source‑based taxation because the taxpayer’s economic connection to the state of residence outweighs the connection between the source state and the immovable property. The connection with the residence state is considered greater, for that state is said to provide the benefits that allow its residents to earn such capital gains (e.g. economy, infrastructure, education, social services). The social and economic environment of the residence state therefore provides the taxpayer with the tools necessary to grow the business carried on in the immovable property located in the source state. In return for these benefits, the taxpayer is expected to pay taxes to the state of residence, not to the source state.

[71]                          In the Minister’s view, this logic holds only where the taxpayer owes economic allegiance to the state of residence, which can be tested by the presence of “sufficient substantive economic connections” to that state. In the absence of such connections, the residence state’s greater claim to tax would retreat so that the connection between the source state and the immovable property would prevail. As a consequence, the source state would possess the greater claim, and the carve-out from source-based taxation should be unavailable.

[72]                          The theory of economic allegiance is indeed the principle underlying the allocation of taxing rights, and giving effect to this principle thus constitutes the broad purpose of provisions of the Treaty, such as art. 13, that distribute the right to tax between the residence and source states (J. Li and F. Avella, “Article 13: Capital Gains”, in Global Tax Treaty Commentaries, last reviewed May 30, 2020 (online), at s. 1.1.2.1). However, I disagree with the Minister’s articulation of the theory in this case.

[73]                          Broadly speaking, the apportionment of taxing rights between the residence and source states under the OECD Model Treaty, which serves as a model for the Treaty, is centred on the distinction between active and passive income (Li and Cockfield, at p. 12; Avi-Yonah, Sartori and Marian, at p. 155). The source state has the primary right to tax active income (e.g. business profits and employment income), and the residence state has only residual rights. Pursuant to the theory of economic allegiance, the source state has a greater claim to tax active income because its economic environment has the closest connection with the origin of wealth (Malherbe, at p. 56; Li and Cockfield, at pp. 66 and 151). Non‑residents owe allegiance to the source state as a result, and they are expected to pay tax for the public services from which they benefit in carrying on their active economic activities in the source state.

[74]                          Conversely, the residence state has the primary right to tax passive income (e.g. interest, dividends, and capital gains), and the source state has only residual rights. The source state’s claim to tax passive income is considered weaker in comparison to that of the residence state because generating such income is assumed to require few public services from the source state. Moreover, the economic environment of the source state is considered less material to the earning prospect of passive investments, as such passive activities may be conducted in various jurisdictions without either improving or negatively affecting their earning prospect. Therefore, non-residents earning passive income owe little allegiance to the source state.

[75]                          The allocation of the right to tax capital gains in the Treaty relies on this articulation of economic allegiance distinguishing between active and passive income. This allocation serves as the broad purpose of art. 13. Under art. 13(5) of the Treaty, the residence state has the primary right to tax capital gains, as they are passive income. Under art. 13(1) to (4), there are exceptions allowing the source state to tax capital gains realized by non-residents. For example, gains derived from the alienation of immovable property, movable property forming part of the business property of a permanent establishment located in the source state, and shares whose value is derived principally from immovable property may be taxed by the source state. The rationale of these exceptions is that the origin of the wealth acquired from sales of immovable property and the like is the source state, which therefore has a greater claim to tax (Malherbe, at pp. 58-59). For instance, the sale of immovable property situated in Canada is, in essence, a sale of a “piece of Canada” — “the ‘Canadianness’ of the property . . . is the source of the gains” (Li and Cockfield, at pp. 198 and 151).

[76]                          The business property exemption applies where a capital gain is realized on the sale of shares whose value is derived principally from immovable property in which a business was carried on. As a consequence, the default rule is reinstated, so that the residence state has the primary right to tax the gain. In my opinion, this constitutes a departure from the theory of economic allegiance as articulated in the Treaty and the OECD Model Treaty and shows that the business property exemption has a different purpose. According to the logic of economic allegiance, the source state normally has a greater claim to tax income derived from a business carried on within its territory or from the disposition of immovable property located within its territory, because the source state’s economic environment has the closest connection to the origin of wealth. Under art. 13(4) and (5), however, the taxing right is allocated to the residence state instead of the source state. This departure can be explained by the fact that economic allegiance is not the sole principle or policy consideration underlying the rules applicable to source-based taxation; the principle of capital import neutrality, the concern to prevent tax base erosion, and the desire to attract foreign investment also underlie these rules (Li and Cockfield, at pp. 151-54). Since all these principles and policy considerations cannot be accommodated in every single rule, a balancing exercise is inevitable (see Shell Canada Ltd. v. Canada, [1999] 3 S.C.R. 622, at para. 43; Canada Trustco, at para. 53; Sun Indalex Finance, LLC v. United Steelworkers, 2013 SCC 6, [2013] 1 S.C.R. 271, at para. 174). Article 13(4) is also the result of a balancing exercise. The theory of economic allegiance is not the dominating rationale underlying the carve-out provided for in art. 13(4). Rather, the main objective is to attract foreign investment, as I explain below.

[77]                          Canada was and still is a large importer of foreign capital and thus a source country (Arnold (2009), at pp. 10-11). As a source country, Canada would have had an interest in negotiating broad source-based taxation rights in bilateral treaties in order to collect larger tax revenues. According to learned authors, Canada must have been alive, however, to the fact that its position on source-based taxation could also be used as an instrument to attract foreign investment crucial to its economy (Li and Cockfield, at p. 153). Harsh source taxes chase away foreign investors, whereas tax breaks attract them. Importers of capital thus have an interest in maintaining a balance between these disincentives and incentives if they want to remain competitive in the global economy. Li and Avella say that Canada has routinely included the carve-out in many of its tax treaties, including this Treaty, for this specific reason (s. 3.1.4.6.3). This tax break encourages foreigners to invest in immovable property situated in Canada in which businesses are carried on (e.g. mines, hotels, or oil shales), rather than simply to invest in assets to be held for speculative purposes.

[78]                          Importantly, I note that only a very small number of the world’s tax treaties include the carve-out, thereby signalling that its inclusion in the Treaty was no accident on Canada’s part. To begin with, less than 35 percent of the world’s pre-2014 double‑tax treaties include a provision allocating to the source state the right to tax capital gains from the disposition of shares of a company or interests in a partnership, trust, or estate whose value is derived principally from immovable property, as provided for in the first part of art. 13(4) (Li and Avella, at ss. 2.1.5 and 3.1.4.1.2). And out of that 35 percent, an even smaller number of treaties contain the carve-out provided for at the end of art. 13(4) (Li and Avella, at s. 3.1.4.6.3).

[79]                          The remaining question is whether the contracting states intended that persons without “sufficient substantive economic connections” to their state of residence be able to take advantage of the carve-out to avoid paying any taxes. Put simply, the question is whether the use of conduit corporations in this context perverts the bargain struck between Canada and Luxembourg. In my view, it does not.

[80]                          The GAAR was enacted to catch unforeseen tax strategies. However, the use of conduit corporations, “legal entit[ies] created in a State essentially to obtain treaty benefits that would not be available directly”, was not an unforeseen tax strategy at the time of the Treaty (“Commentary on Article 1” of the 1998 OECD Model Treaty, at para. 9). Indeed, it was far from being a novel phenomenon that emerged subsequently to the signing of the Treaty. Back in the 1970s, the “Commentary on Article 1” of the 1977 OECD Model Treaty criticized tax planning strategies involving conduit corporations (paras. 8-9). This was also an issue discussed by learned authors (see, e.g., A. A. Knechtle, Basic Problems in International Fiscal Law (1979), at p. 115; D. R. Davies, Principles of International Double Taxation Relief (1985), at paras. 1.13 and 3.22). And more contemporaneously to the Treaty, the “Commentary on Article 1” of the 1998 OECD Model Treaty discussed the use of conduit corporations as well.

[81]                          Luxembourg is a country well known for its broad tax treaty network and international tax haven regime, making it an attractive jurisdiction to set up a conduit corporation and take advantage of treaty benefits. As mentioned above, one can presume that Canada had knowledge of these features of Luxembourg’s tax system when it entered into the Treaty. Canada nevertheless entered into a bilateral tax treaty with Luxembourg with only minimal safeguards and thereby ignored many of the OECD’s suggestions. At that time, as discussed above, the “Commentary on Article 1” of the 1998 OECD Model Treaty set out a whole menu of potential anti-avoidance provisions that might have short-circuited the creation of conduit corporations in Luxembourg.

[82]                          I acknowledge that the absence of specific anti-avoidance rules that would have prevented the situation is not necessarily determinative of the application of the GAAR (see Copthorne, at paras. 108-11). Of course, one could always imagine a potential anti-avoidance rule that would have pre-empted the tax strategy at issue. If that were the standard, I agree that it would provide a full response in every case and gut the GAAR. In this case, the absence of specific anti-avoidance provisions represents, however, an enlightening contextual and purposive element as it sheds light on the contracting states’ intention. This is not a case where Parliament did not or could not have foreseen the tax strategy employed by the taxpayer. Options to remediate the situation were available and known by the parties, but they made deliberate choices to guard some benefits against conduit corporations and to leave others unguarded. Had the parties truly intended to prevent such corporations from taking advantage of the carve-out, they could have done so. Combined with Canada’s preference at the time of the Treaty for taking advantage of economic benefits yielded by foreign investments rather than higher tax revenues (as will be discussed below), this makes the rationale of the carve-out even clearer. In my opinion, Canada and Luxembourg made a deliberate choice to leave the business property exemption unguarded.

[83]                          The parties agreed to exclude Luxembourg holding companies from their Treaty (art. 28(3)). However, as explained above, the parties did not follow the OECD’s suggestion to include a “look-through” provision combined with a provision safeguarding bona fide business activities. Doing so would have excluded conduit corporations that were owned by residents of a third country and that conducted few “substantive business activities” in Luxembourg.

[84]                          Moreover, Luxembourg and Canada added provisions reserving the benefits of the Treaty to the beneficial owners of certain income, but only in respect of dividends, interest, and royalties, not capital gains (arts. 10 to 12). If the parties had applied the concept of beneficial ownership to the carve-out, it would have prevented conduit corporations from taking advantage of this benefit where their beneficial owners were residents of a third country (see, e.g., Prévost Car). Although the OECD Model Treaty may not have specifically recommended extending the concept of beneficial ownership to capital gains, nothing barred the parties from doing so. After all, the OECD Model Treaty remains a model, not a binding legal instrument. I fail to see, as Rowe and Martin JJ. do, a fundamental difference between capital gains, on the one hand, and dividends, interest, and royalties, on the other, that would have made beneficial ownership unsuitable in respect of capital gains. It is true that, in a paper prepared in 2008 for the United Nations, Professor Philip Baker questioned the possible extension of the beneficial ownership concept to capital gains because it is harder to identify the beneficial owner of a capital gain than to identify the beneficial owner of a traceable flow of income like dividends, interest, and royalties (The United Nations Model Double Taxation Convention Between Developed and Developing Countries: Possible Extension of the Beneficial Ownership Concept, U.N. Doc. E/C.18/2008/CRP.2/Add.1, Ann., October 17, 2008, at para. 55). But Professor Baker explained that this difference is “not insurmountable” (para. 58). In the end, he even suggested the potential wording of a beneficial ownership provision applicable to capital gains:

6. Gains from the alienation of any property other than that referred to in paragraphs 1, 2, 3, 4 and 5 shall be taxable only in the Contracting State of which the beneficial owner is a resident. [para. 59]

[85]                          As suggested in the “Commentary on Article 13” of the 1998 OECD Model Treaty, Canada could also have insisted on a subject-to-tax provision if it had truly been concerned about the erosion of its tax base (para. 21). Pursuant to such a provision, a contracting state foregoes its right to tax capital gains only if the other state actually taxes these gains. The inclusion of a provision of this sort would have meant that capital gains exempt from or subject to very little tax in Luxembourg would be taxable in Canada, instead of having almost double non-taxation. Canada’s decision is, however, not surprising. Situations of double non-taxation are intentional and part of the bargain in most cases (Couzin, at p. 109). The absence of a subject-to-tax provision, combined with Canada’s knowledge of Luxembourg’s tax system, confirms my view that Canada’s primary objective in including art. 13(4) was to cede its right to tax capital gains of a certain nature realized in Canada in order to attract foreign investment. It was not part of the bargain that Luxembourg actually tax the gains to the same extent that Canada would have taxed them.

[86]                          Further, consideration of tax treaties in pari materia forming part of the broader context of the Treaty corroborates the parties’ deliberate choice to leave the business property exemption unguarded against conduit corporations established in Luxembourg. As noted by Professor Arnold, Canada began including discrete purpose tests designed to prevent treaty shopping in some of the treaties it entered into around the same time it entered into the Treaty, which was signed in 1999 ((2009), at p. 358). See, e.g., the treaties with Nigeria^[2], Ukraine^[3], Kazakhstan^[4], Uzbekistan^[5], and Peru^[6]. These provisions were designed to take back certain treaty benefits where the purpose of a transaction was to gain access to these benefits. Most likely, the creation of a conduit corporation in Luxembourg to access the exemption would have been caught by such a purpose test. Had the parties truly intended to deprive such corporations of the benefits of the carve-out, they would have made the carve-out subject to a purpose test.

[87]                          The absence of any such anti-avoidance measure that would have limited access to the carve-out in a treaty with a country known for not taxing capital gains leads me to believe that Canada weighed the pros and cons and concluded that its national interest in attracting foreign investors, using Luxembourg as a conduit to take advantage of the carve-out, outweighed its interest in collecting more tax revenues on such capital gains. This answers the question of “why the benefit was conferred” posed under the GAAR (Canada Trustco, at para. 66). This choice must also have been motivated by the fact that Canada was not keen on going its own way at a time when the international community was not yet as serious about curtailing treaty shopping as it was during the years leading to the signature and ratification of the Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting, Can. T.S. 2019 No. 26, in 2017 and as it has remained to this day (see Multilateral Instrument in Respect of Tax Conventions Act, S.C. 2019, c. 12; Arnold (2009), at p. 18). As a relatively small country, “Canada does not have the luxury of setting its own policy without considering what other countries do”, and it must have rightly seen multilateralism as the way forward (Li and Cockfield, at p. 25).

[88]                          This is not an absurd proposition, as my colleagues assert. There is no better way to describe Canada’s attitude at the time of the Treaty when faced with the dilemma between higher tax revenues and competitiveness than to quote the words of the Department of Finance’s response to the Report of the Auditor General of Canada to the House of Commons, 1992, at p. 52:

To a large degree, international norms limit the range of options available to the Canadian government and, in this context, the government’s policy has generally been to favour competitiveness concerns over those of revenue generation. [Emphasis added.]

[89]                          In conclusion, the object, spirit, and purpose of the carve-out provided for in art. 13(4) and (5) of the Treaty are to foster international investment by exempting residents of a contracting state from taxes in the source state on capital gains realized on the disposition of immovable property in which a business was carried on, or on the disposition of shares whose value is derived principally from such immovable property. The fact that the capital gains may not be taxed in Luxembourg, leading to double non‑taxation, and the fact that conduit corporations can take advantage of the carve-out are tax planning outcomes consistent with the bargain struck between Canada and Luxembourg. Although Canada had a greater claim to tax such income based on the theory of economic allegiance and would have most likely received more tax revenues without the carve-out given its traditional status as a source state, Canada agreed to forego its right to tax such capital gains, regardless of whether they would be taxed in Luxembourg, in order to attract foreign investment in business assets embodied in immovable property located in Canada (e.g. hotels, mines, or oil shales) and to reap the economic benefits generated by that investment. This is what was actually agreed upon. It appears that Canada was aware of tax planning strategies using conduit corporations in Luxembourg but that it made a deliberate choice not to include anti-avoidance provisions that would have addressed this situation in the Treaty.

E.            Second Step: Abusiveness of the Transaction

[90]                          The second step of the methodology for the abuse analysis is “to examine the factual context of a case to determine whether the avoidance transaction defeated or frustrated the object, spirit or purpose of the provisions in issue” (Canada Trustco, at para. 55).

[91]                          Alta Luxembourg was created as part of a restructuring of activities to take advantage of the carve-out provided for in art. 13(4) of the Treaty. It concedes that this tax benefit was derived from a transaction not arranged primarily for a bona fide purpose other than to obtain this benefit. However, I fail to see how this avoidance transaction was abusive.

[92]                          In this Court, as in the Federal Court of Appeal, the Minister rightly concedes that Alta Luxembourg is a resident of Luxembourg for the purposes of the Treaty, as it has its legal seat there. Moreover, nothing indicates that Alta Luxembourg was not liable to tax in Luxembourg. The gain it realized on the disposition of its shares of Alta Canada was reported to the Luxembourg authorities and was subject to full taxation under their domestic laws. The fact that Alta Luxembourg paid less tax in Luxembourg on this gain than it would have paid if it were a Canadian resident paying tax in Canada changes nothing in this determination. As noted by this Court in Copthorne, “determining the rationale of the relevant provisions of the Act should not be conflated with a value judgment of what is right or wrong nor with theories about what tax law ought to be or ought to do” (para. 70). Therefore, from the moment that Alta Luxembourg realized a gain on the disposition of its shares of Alta Canada, the laws of Luxembourg applied and Canada’s interest in the gain ceased. As a result, Alta Luxembourg’s resident status falls squarely within the object, spirit, and purpose of arts. 1 and 4(1) of the Treaty.

[93]                          The Minister has failed to discharge her burden of establishing that the avoidance transaction defeats the underlying rationale, the “object, spirit, or purpose”, beyond the words of the provisions. Alta Luxembourg met the clear requirements of arts. 1, 4, and 13(4).

[94]                          The Treaty makes it clear that Canada and Luxembourg agreed that the power to tax would be allocated to Luxembourg where the conditions of the carve-out were met. There is nothing in the Treaty suggesting that a single‑purpose conduit corporation resident in Luxembourg cannot avail itself of the benefits of the Treaty or should be denied these benefits due to some other consideration such as its shareholders not being themselves residents of Luxembourg. In this case, the provisions operated as they were intended to operate; there was no abuse, and, therefore, the GAAR cannot be applied to deny the tax benefit claimed.

[95]                          This result accords with the true intention of the partners to the Treaty and must be respected. As Pelletier J.A. held in MIL (Investments), “[t]o the extent that the appellant argues that the Tax Treaty should not be interpreted so as to permit double non-taxation, the issue raised by GAAR is the incidence of Canadian taxation, not the foregoing of revenues by the Luxembourg fiscal authorities” (para. 8).

[96]                          A final note on the Minister’s implication that treaty shopping arrangements are inherently abusive. A broad assertion of “treaty shopping” does not conform to a proper GAAR analysis. In accordance with the separation of powers, developing tax policy is the task of the executive and legislative branches. Courts do not have the constitutional legitimacy and resources to be tax policy makers (Canada Trustco, at para. 41). It is for the executive and legislative branches to decide what is right and what is wrong, and then to translate these decisions into legislation that courts can apply. It bears repeating that the application of the GAAR must not be premised on “a value judgment of what is right or wrong [or] theories about what tax law ought to be or ought to do” (Copthorne, at para. 70). Taxpayers are “entitled to select courses of action or enter into transactions that will minimize their tax liability” (Copthorne, at para. 65). The courts’ role is limited to determining whether a transaction abuses the object, spirit, and purpose of the specific provisions relied on by the taxpayer. It is not to rewrite tax statutes and tax treaties to prevent treaty shopping when these instruments do not clearly do so.

VI.         Conclusion

[97]                          For these reasons, I would dismiss the appeal. The respondent is entitled to its costs in this appeal.

 

The reasons of Wagner C.J. and Rowe and Martin JJ. were delivered by

 

                    Rowe and Martin JJ. —

I.               Overview

[98]                         Multinational companies exploiting gaps and mismatches in international tax rules erode domestic tax bases and cost countries an estimated US$100 to US$240 billion in lost revenue annually (Organisation for Economic Co‑operation and Development (“OECD”), Background Brief: Inclusive Framework on BEPS, January 2017 (online), at p. 9). As a result, aggressive international tax avoidance has received increasing political and media scrutiny (see, e.g., The Economist’s “The big carve-up” (May 15, 2021), at pp. 65-66, and Le Monde’s series OpenLux: Enquête sur le Luxembourg, coffre-fort de l’Europe by J. Baruch et al., February 8, 2021 (online)).

[99]                         Although it is a long standing principle in Canadian law that taxpayers may arrange their affairs to minimize their amount of tax payable (Commissioners of Inland Revenue v. Duke of Westminster, [1936] A.C. 1 (H.L.); Canada Trustco Mortgage Co. v. Canada, 2005 SCC 54, [2005] 2 S.C.R. 601, at para. 11), the freedom to do so is not without limits. Canada has acted to curb abusive international tax avoidance by enacting the general anti-avoidance rule (“GAAR”), which denies tax benefits when taxpayers engage in transactions that conform with the text of the tax rules relied upon, but do not accord with their rationale. In introducing the GAAR in our tax legislation some 30 years ago, and in making it clear that it was applicable to abuses of tax treaties, Parliament made a policy choice by which it intended to fight harmful tax avoidance schemes that cross the line of legitimate tax planning and venture into the realm of abusive tax avoidance.

[100]                     Courts now have the responsibility to give proper effect to the intention of Parliament and ensure the GAAR plays a meaningful role in controlling avoidance transactions that technically comply with the provisions of a tax treaty but frustrate their underlying rationale. The interpretation exercise that is mandated in a GAAR analysis thus vests upon courts the unusual duty to look beyond the words of the applicable provisions to determine whether the transactions in question frustrate the underlying rationale of those provisions. Giving an interpretation confined to the black letter of these legislative provisions instead defeats Parliament’s will and fails to fulfil the courts’ role “to interpret and apply the Act as it was adopted by Parliament” (Shell Canada Ltd. v. Canada, [1999] 3 S.C.R. 622, at para. 45; see also Canada Trustco, at para. 13).

[101]                     Given that the GAAR can only find application where a taxpayer has complied with the strict requirements of a provision, absolute certainty cannot be achieved, nor was it intended. This is a legislative choice that Parliament made in order to strike a necessary balance between the uncertainty inherent in the GAAR and the fairness of the Canadian tax system as a whole achieved by defeating abusive tax avoidance schemes.

[102]                     Alta Energy Luxembourg S.A.R.L. (“Alta Luxembourg”) is a Luxembourg corporation created so that shareholders of Alta Energy Partners, LLC (“Alta US”), who are residents neither of Canada nor of Luxembourg, could benefit from an exemption from Canadian income tax under the Convention between the Government of Canada and the Government of the Grand Duchy of Luxembourg for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and on Capital, Can. T.S. 2000 No. 22 (“Treaty”), with respect to liquidation of an investment in oil and gas properties. Despite being a mere conduit interposed in Luxembourg, Alta Luxembourg claimed a tax exemption for a capital gain of more than $380 million. The issue in this case is whether this transaction constitutes abuse under the GAAR. In effect, does Alta Luxembourg get to claim this exemption and, thereby, pay no tax at all in Canada on a multimillion dollar profit made from property in Canada, even though it had no genuine connection to Luxembourg?

[103]                     We conclude that Alta Luxembourg’s claim for a tax benefit under the Treaty is the result of abusive avoidance transactions. In our respectful view, the courts below did not properly identify the rationale underlying the relevant provisions of the Treaty. They gave weight only to the text and failed to consider why the provisions were put in place. This is not the exercise mandated under the GAAR.

[104]                     The object, spirit or purpose of the relevant provisions of the Treaty is to assign taxing rights to the state with the closest economic connection to the taxpayer’s income. Article 13(4) allocates to Luxembourg the right to tax its residents’ indirect gains from immovable property situated in Canada used in a business. In such cases, Luxembourg is deemed to have the closer economic connection with the taxpayer’s income because the business activity, rather than the immovable property itself, drives the value of the property.

[105]                     Here, a review of the factual matrix reveals that Alta Luxembourg utterly lacks a genuine economic connection with Luxembourg. Allowing it to benefit from art. 13(4) and (5) would frustrate the rationale of this provision and is thereby abusive. The common intention of Canada and Luxembourg in entering into a tax treaty could not have been to provide avenues for residents of third-party states to indirectly obtain benefits from Luxembourg’s tax regime they could not obtain directly, despite the absence of genuine ties to that state. Accordingly, we would allow the appeal.

II.            Background

[106]                     In June 2011, Alta Energy Partners Canada Ltd. (“Alta Canada”) was incorporated under the laws of Alberta. It was a wholly owned subsidiary of Alta US, whose shareholders were foreign investors from the United States or other countries. Alta Canada carried on an unconventional shale oil business in Northern Alberta. The indirect owners of the shares of Alta Canada were advised by tax professionals that their current corporate structure was undesirable from a tax perspective — the capital gain resulting from the future disposition of the shares would be subject to Canadian taxes because the shares of Alta Canada were taxable Canadian property and Canada reserves the right to tax such capital gains under the Convention Between Canada and the United States of America with Respect to Taxes on Income and on Capital (enacted in law in Canada by the Canada-United States Tax Convention Act, 1984, S.C. 1984, c. 20, Sch. I; see ss. 2(3)(c) and 115(1)(b) of the Income Tax Act, R.S.C. 1985, c. 1 (5th Supp.) (“Act”). The advice highlighted the existence of a favourable exemption under other tax treaties that would lead to the non-taxation of the capital gain to be realized on the sale of the shares.

[107]                     In April 2012, Alta Luxembourg was formed under the laws of Luxembourg for the purpose of taking advantage of the Treaty. The shares of Alta Luxembourg were still indirectly controlled by shareholders of Alta US. Alta Luxembourg had no genuine connection to Luxembourg: it had no business activity in Luxembourg and it did not hold any other investment. The sole purpose for the incorporation was to avoid paying taxes in Canada on the sale of the shares of Alta Canada.

[108]                     On the same day, Alta US sold all its shares of Alta Canada to Alta Luxembourg. The investors had received a confirmation from the Luxembourg tax authorities that capital gains realized on the participation in Alta Canada would be exempt from Luxembourg tax. The result was to replace Alta US with Alta Luxembourg as shareholder of Alta Canada, but to keep the same persons effectively in control.

[109]                     In 2013, Alta Luxembourg sold the shares of Alta Canada for an amount of approximately $680 million. The resulting capital gain was in excess of $380 million. Alta Luxembourg then claimed an exemption from Canadian income tax under art. 13(5) of the Treaty.

[110]                     The Minister of National Revenue denied the exemption, partially on the basis that the GAAR operated to deny the benefit. Alta Luxembourg appealed the assessment. The Tax Court of Canada allowed Alta Luxembourg’s appeal. It found that Alta Luxembourg could claim the exemption from Canadian income tax and that the GAAR did not prevent this entitlement (2018 TCC 152, [2019] 5 C.T.C. 2183). The Federal Court of Appeal agreed with the Tax Court (2020 FCA 43, [2020] 5 C.T.C. 193).

III.         Analysis

A.           The Legal Framework Under the General Anti-Avoidance Rule

[111]                     The GAAR, set out at s. 245 of the Act, is a tool to preserve the integrity of our tax system which seeks to respond to “the need to prevent further erosion of tax revenues through purely tax-motivated strategies” (J. Sasseville, “Implementation of the General Anti-Avoidance Rule”, in Corporate Management Tax Conference, Income Tax Enforcement, Compliance, and Administration (1988), 4:1, at p. 4:2). Indeed, an unbridled application of the Duke of Westminster principle providing that a taxpayer may arrange their affairs to minimize their tax burden “can mislead taxpayers into believing that tax plans that merely comply with the technical provisions of the Act are acceptable” (V. Krishna, Income Tax Law (2nd ed. 2012), at p. 473).

[112]                     Establishing the line between legitimate tax planning and abusive tax avoidance is a complex balancing exercise. Parliament drew that line by enacting the GAAR — “the apex of all anti-avoidance measures” (Krishna (2012), at p. 470). Section 245, which sets out the GAAR reads, in part:

(2) Where a transaction is an avoidance transaction, the tax consequences to a person shall be determined as is reasonable in the circumstances in order to deny a tax benefit that, but for this section, would result, directly or indirectly, from that transaction or from a series of transactions that includes that transaction.

 

. . .

 

(4) Subsection (2) applies to a transaction only if it may reasonably be considered that the transaction

 

(a) would, if this Act were read without reference to this section, result directly or indirectly in a misuse of the provisions of any one or more of

 

. . .

 

(iv) a tax treaty . . . or

 

. . .

 

(b) would result directly or indirectly in an abuse having regard to those provisions, other than this section, read as a whole.

[113]                     There are three requirements for the GAAR to apply and to deny tax benefits: (1) there must be a tax benefit; (2) the transaction giving rise to the tax benefit must be an avoidance transaction under s. 245(3), “in the sense that it cannot be said to have been reasonably undertaken or arranged primarily for a bona fide purpose other than to obtain a tax benefit”; and (3) the avoidance transaction must result in a misuse or abuse of the Act or a tax treaty, “in the sense that it cannot be reasonably concluded that a tax benefit would be consistent with the object, spirit or purpose of the provisions relied upon by the taxpayer” (Canada Trustco, at para. 66(1.); s. 245(3) of the Act). Thus, where there is a tax benefit and an avoidance transaction, the application of the GAAR to deny the tax benefit will only be avoided “if it may be reasonable to consider that it did not result from abusive tax avoidance under s. 245(4)” (para. 35, see also Department of Finance, Explanatory Notes to Legislation Relating to Income Tax (1988), at p. 464). The terms “misuse” and “abuse” are not defined in the Act. As such, courts had to ascertain the appropriate test for determining whether abusive tax avoidance had taken place.

(1)          Determining Whether There Has Been Abuse Under the GAAR

[114]                     The third requirement of the GAAR is subdivided into two steps. The first is to determine the object, spirit or purpose of the provisions giving rise to the tax benefit; the second is to determine whether the transaction frustrates that purpose (Canada Trustco, at para. 44).

[115]                     First, the court must identify the “object, spirit or purpose of the provisions of the Income Tax Act that are relied on for the tax benefit, having regard to the scheme of the Act, the relevant provisions and permissible extrinsic aids” (Canada Trustco, at para. 55). As Justice Rothstein emphasized in Copthorne Holdings Ltd. v. Canada, 2011 SCC 63, [2011] 3 S.C.R. 721, Parliament, in enacting the GAAR, has conferred an “unusual duty” on the courts to look beyond the words of the statute to ascertain the object, spirit or purpose of the provisions at issue (para. 66). Even if the taxpayer complies with the text of the provisions of the Act or tax treaty, courts must determine whether the taxpayer’s transactions are in accord with the rationale of the provisions relied upon (Copthorne, at para. 66; Krishna (2012), at p. 486).

[116]                     Although the GAAR analysis involves a textual, contextual, and purposive analysis, the inquiry differs from traditional statutory interpretation. Whereas the object of traditional statutory interpretation is to “determine what the words of the statute mean”, the GAAR analysis is a “search . . . for the rationale that underlies the words that may not be captured by the bare meaning of the words themselves” (Copthorne, at para. 70). The question under the GAAR is not whether the taxpayer can claim a tax benefit, but rather “why the benefit was conferred” (Canada Trustco, at para. 66(4.)). Thus, a traditional statutory interpretation “is constrained by the text of the relevant provisions in a way that a GAAR analysis is not” (D. G. Duff, “The Interpretive Exercise Under the General Anti-Avoidance Rule”, in B. J. Arnold, ed., The General Anti-Avoidance Rule — Past, Present, and Future (2021), 383, at p. 406). This follows by necessary implication. To determine whether a taxpayer can claim a tax benefit in the first place, the provisions of the Act must be interpreted applying the traditional textual, contextual and purposive approach. And, in any GAAR case, “the text of the provisions at issue will not literally preclude a tax benefit the taxpayer seeks by entering into the transaction or series” (Copthorne, at para. 88). If the GAAR simply replicated the statutory interpretation exercise, it would be meaningless. This difference in the methodology of interpretation is pivotal to giving proper effect to the GAAR.

[117]                     Similarly, in determining the meaning of tax treaties, art. 31 of the Vienna Convention on the Law of Treaties, Can. T.S. 1980 No. 37, requires the terms of a treaty to be interpreted in light of their context and in light of the treaty’s object and purpose. The GAAR directs courts to go behind this meaning and to identify the rationale underlying the relevant provision of the treaty.

[118]                     Once the court has identified the rationale underlying the relevant provisions, the second step of the abuse analysis is “to determine whether the avoidance transaction defeated or frustrated the object, spirit or purpose of the provisions in issue” (Canada Trustco, at para. 55). There will be a finding of abuse: “. . . (1) where the transaction achieves an outcome the statutory provision was intended to prevent; (2) where the transaction defeats the underlying rationale of the provision; or (3) where the transaction circumvents the provision in a manner that frustrates or defeats its object, spirit or purpose . . .” (Copthorne, at para. 72). These considerations “are not independent of one another and may overlap” (ibid.). The abusive nature of the transaction “must be clear” (Canada Trustco, at para. 62).

[119]                     In summary, the GAAR will not apply “to deny the tax benefi[t] that results from these transactions as long as they are carried out within the object and spirit of the provisions of the Act read as a whole” (Krishna (2012), at p. 486, quoting Department of Finance (1988), at p. 464). The same rule would undoubtedly apply to tax benefits arising under the Treaty. However, the GAAR applies when the taxpayer carries out transactions to obtain a tax benefit not intended by specific provisions of the Act or treaty read as a whole. In such cases, the GAAR “will override other provisions of the Act [or tax treaties] since, otherwise, its object and purpose would be defeated” (p. 489, quoting Department of Finance (1988), at p. 465).

(2)          Certainty, Predictability and Fairness

[120]                     To the extent that the GAAR is only invoked where there will be compliance with the requirements on a literal reading of a provision, the GAAR, by nature, gives rise to an “unavoidable degree of uncertainty for taxpayers” (Copthorne, at para. 123). Allowing the GAAR to create this uncertainty was a deliberate choice that Parliament made when it enacted a provision that can defeat tax avoidance schemes that exploit Canada’s legislation and treaties. The Act contains numerous anti-avoidance provisions that target specific factual scenarios. Inventive tax planners may successfully avoid these provisions by using structured transactions, but those transactions may still constitute abusive tax avoidance. The GAAR was designed to act as the final, catch-all provision which can be applied where no specific anti-avoidance provision applies, exists or was circumvented. It is a “provision of last resort” (Canada Trustco, at para. 21).

[121]                       Absolute certainty under the GAAR does not exist. Parliament has made a tax policy choice in order to balance the principles of certainty and predictability with another principle that is as fundamental to our tax system — the principle of fairness (Lipson v. Canada, 2009 SCC 1, [2009] 1 S.C.R. 3, at para. 52). The GAAR “is designed, in the complex context of the [Act], to restrain abusive tax avoidance and to make sure that the fairness of the tax system is preserved” (ibid.). Sophisticated taxpayers who can afford tax professionals have access to planning strategies that lower or eliminate their tax burden through what may cross the line into abusive tax avoidance territory. Not applying the GAAR to those abusive schemes is deeply unfair not solely because only this select group of taxpayers may have access to such professionals, but also because the tax burden avoided by the select group falls back on the taxpayers who do not, for instance through higher tax rates (D. A. Dodge, “A New and More Coherent Approach to Tax Avoidance” (1988), 36 Can. Tax J. 1, at pp.  4, 8 and 9).

[122]                       Professor J. Li underscores this essential balancing role the GAAR plays in our tax system, as recognized in our jurisprudence:

As the Supreme Court noted in Canada Trustco, the preservation of “certainty, predictability and fairness” for individual taxpayers is considered a “basic tenet of tax law” [at para. 61]. On the other hand, the GAAR requires the balancing of this set of policy concerns against the concern for protection of the tax base and the fairness of the tax system as a whole.

 

. . . the GAAR cases generally involve situations that do not concern the majority of taxpayers, and the transactions are well planned and executed on the basis of professional tax advice. Therefore, the requirement of certainty does not ring true in GAAR cases. [Emphasis added; footnotes omitted.]

 

(“‘Economic Substance’: Drawing the Line Between Legitimate Tax Minimization and Abusive Tax Avoidance” (2006), 54 Can. Tax J. 23, at p. 40)

[123]                     Consequently, a finding that the GAAR applies to deny the tax benefits conferred by clear provisions where avoidance transactions defeat their underlying rationale does not run counter to the principles of certainty, predictability and fairness. The application of the GAAR in these circumstances upholds the balance Parliament sought to strike between those principles and gives effect to its intent to curb abusive tax avoidance (Department of Finance (1988), at p. 461, cited in Canada Trustco, at para. 15).

[124]                     This is the view that the majority upheld in Lipson. In that case, the dissenting judges were of the view that the Duke of Westminster principle should prevail over the GAAR on the basis that, if not contained, the GAAR was a “weapon” that “could have a widespread, serious and unpredictable effect on legitimate tax planning” (para. 55). Justice LeBel, for the majority, found that this approach “essentially gut[ted] the GAAR and rea[d] it out of the [Act] under the guise of an exercise in legal interpretation” (para. 52). Overemphasizing the principles of certainty and predictability to the detriment of that of fairness is contrary to Parliament’s will. In the words of LeBel J., the “desire to avoid uncertainty cannot justify ignoring a provision of the [Act] that is clearly intended to apply to transactions that would otherwise be valid on their face” (ibid.). Courts have the obligation to animate all parts of Parliament’s legislation, including s. 245 of the Act.

(3)          The Application of the GAAR to Bilateral Tax Conventions

[125]                     Following the introduction of s. 245 of the Act, the question of whether this provision applied to tax treaties gave rise to much debate among commentators and scholars (W. I. Innes, P. J. Boyle and J. A. Nitikman, The Essential GAAR Manual: Policies, Principles and Procedures (2006), at p. 198-204). This debate is now settled.

[126]                     As  Bowman J. (as he then was) had opined in RMM Canadian Enterprises Inc. v. R., [1998] 1 C.T.C. 2300, at paras. 52-54 and 56, quoting McNichol v. The Queen (1997), 97 D.T.C. 111, at pp. 120-21, the Minister could use the GAAR to deny tax benefits arising from the exploitation of Canada’s tax treaties. Parliament later confirmed that s. 245 applied to tax treaty benefits when it enacted s. 245(4)(a)(iv) of the Act in the Budget Implementation Act, 2004, No. 2, S.C. 2005, c. 19, s. 52(2). It also enacted s. 4.1 of the Income Tax Conventions Interpretation Act, R.S.C. 1985, c. I-4 (“ITCIA”), which provides that “[n]otwithstanding the provisions of a convention . . . it is hereby declared . . . that section 245 of the [Act] applies to any benefit provided under the convention.” And, in the event of inconsistency between the Treaty and the ITCIA, the provisions of that act prevail (Income Tax Conventions Implementation Act, 1999, S.C. 2000, c. 11, s. 51(2)). The courts are bound to give effect to Parliament’s legislative efforts to clearly communicate its intent to curb tax avoidance arising from an improper use of some of Canada’s tax treaties through the application of the GAAR. The result is that “tax treaties must be interpreted in the same manner as domestic legislation when analyzing potentially abusive avoidance transactions” (MIL (Investments) S.A. v. R., 2006 TCC 460, [2006] 5 C.T.C. 2552 (“MIL (TCC)”), at para. 28). Contrary to our colleague Côté J.’s suggestion, this is not an impermissible extension of the GAAR. If there were any doubts about the application of the GAAR to tax treaties, they were unequivocally resolved when Parliament amended the Act in 2005 (see Standing Senate Committee on National Finance, Proceedings of the Standing Senate Committee on National Finance, Nos. 19 and 20, 1st Sess., 38th Parl., April 20 and May 2, 2005, at pp. 19:16-17, 19:22, and 19:26-27).

[127]                     Therefore, although this Court’s pronouncements on the GAAR were made in the context of alleged abuses of the Act, there is no dispute that the GAAR also applies to avoidance transactions that abuse or misuse tax treaties.

[128]                     The backdrop against which this amendment was brought provides insight. It followed a report by the Auditor General of Canada on what was then the Canada Customs and Revenue Agency (“CCRA”), in which it identified a number of tax avoidance schemes that exploited some of Canada’s tax treaties. In discussing the tax treaty Canada concluded with Barbados, the Auditor General observed the following:

During our review of non-resident files, we noted that the Agency has discovered a number of schemes developed to exploit the Canada-Barbados Income Tax Agreement. Canada usually signs tax treaties to avoid double taxation. Barbados does not tax capital gains. However, the Agreement allows a resident of Barbados to claim a Canadian tax exemption on a capital gain that would otherwise be subject to Canadian tax. The Agency is currently reviewing transactions that move capital gains from Canada to Barbados. [Emphasis added.]

 

(Office of the Auditor General, Report of the Auditor General of Canada to the House of Commons — Chapter 7 — Canada Customs and Revenue Agency, International Tax Administration: Non-Residents Subject to Canadian Income Tax (2001), at para. 7.85)

[129]                     One of the schemes given as an example of an exploitation of Canada’s tax treaties is strikingly similar to the facts of the present case: a non-resident company held shares of a Canadian company, the proceeds of the sale of which shares would be taxable in Canada. The non-resident company shifted its residence to Barbados and claimed a Canadian tax exemption on the capital gain provided under the treaty between Canada and Barbados (para 7.89).

[130]                     The Auditor General recommended that the CCRA “continue to be vigilant in ensuring that tax treaties are not used inappropriately to reduce Canadian tax and, if necessary, should seek legislative or treaty changes to protect Canada’s tax base” (para. 7.91 (emphasis added)). Parliament opted to make clear that the GAAR would be its preferred tool to curb tax avoidance arising from an improper use of Canada’s tax treaties.

B.            Application

[131]                     We now turn to applying the principles laid out above to the facts of this case. Since the first two steps of the GAAR framework are not at issue, we will focus solely on the abuse analysis.

(1)          Object, Spirit or Purpose of Articles 1, 4 and 13 of the Treaty

[132]                     The Crown submits that, with respect to “shares which derive their value principally from immovable property used in a company’s business”, the rationale underlying the relevant provisions of the Treaty is to allocate taxing rights to the state of residence, as the economic connection to the company’s state of residence would usually outweigh the connection to the source state as the situs of the immovable property (A.F., at para. 65). Alta Luxembourg adopts the view of the courts below, according to which the rationale is no broader than the text of the relevant provisions of the Treaty (A.F., at para. 97). The rationale is simply to “exempt residents of Luxembourg from Canadian taxation where there is an investment in immovable property used in a business” (T.C.C. reasons, at para. 100; see also C.A. reasons, at para. 67).

[133]                     We agree with the Crown. Contrary to the result reached in the courts below, this is not a case where the text of a provision fully expresses its underlying rationale; failing to have regard to this rationale renders the GAAR “meaningless” (Copthorne, at paras. 110-11). We conclude that the rationale underlying arts. 1, 4 and 13 of the Treaty is to assign taxing rights to the state with the stronger economic claim to the income. Canada is presumed to have the stronger connection to capital gains related to passive investment in immovable property situated in Canada realized by residents of Luxembourg. However, when the value of the immovable is driven by a company’s business — that is, when residents of Luxembourg realize a gain from immovable property used in a business — Luxembourg is deemed to have the closer economic connection. The rationale underlying the carve-out in art. 13(4) is to reflect the state of residence’s stronger economic claim to tax the income. In this case, the state of residence is Luxembourg but the state with the stronger economic claim to tax the income is Canada.

(a)           The Text of the Relevant Provisions

[134]                     The relevant provisions of the Treaty are the following:

ARTICLE 1

 

I. Scope of the Convention

 

Persons Covered

 

This Convention shall apply to persons who are residents of one or both of the Contracting States.

 

ARTICLE 4

 

Resident

 

1.  For the purposes of this Convention, the term “resident of a Contracting State” means any person who, under the laws of that State, is liable to tax therein by reason of that person’s domicile, residence, place of management or any other criterion of a similar nature. This term also includes a Contracting State or a political subdivision or local authority thereof or any agency or instrumentality of any such State, subdivision or authority. This term, however, does not include any person who is liable to tax in that State in respect only of income from sources in that State.

 

ARTICLE 13

 

Capital Gains

 

1. Gains derived by a resident of a Contracting State from the alienation of immovable property situated in the other Contracting State may be taxed in that other State.

 

4.  Gains derived by a resident of a Contracting State from the alienation of:

a.  shares (other than shares listed on an approved stock exchange in the other Contracting State) forming part of a substantial interest in the capital stock of a company the value of which shares is derived principally from immovable property situated in that other State; or

b.  an interest in a partnership, trust or estate, the value of which is derived principally from immovable property situated in that other State,

 

may be taxed in that other State. For the purposes of this paragraph, the term “immovable property” does not include property (other than rental property) in which the business of the company, partnership, trust or estate was carried on; and a substantial interest exists when the resident and persons related thereto own 10 per cent or more of the shares of any class or the capital stock of a company.

 

5.  Gains from the alienation of any property, other than that referred to in paragraphs 1 to 4 shall be taxable only in the Contracting State of which the alienator is a resident.

[135]                     Article 1 provides that the Treaty applies to “residents” of Canada or Luxembourg. Article 4 defines “resident” as persons who are liable to pay tax under the laws of the state. Article 13 of the Treaty provides for the allocation of the right to tax capital gains between Canada and Luxembourg. Under art. 13(5), the state of residence (here, Luxembourg) retains its jurisdiction to tax capital gains unless the exceptions in art. 13(1) to (4) of the Treaty apply.

[136]                     Article 13(1) preserves the right of the source state (here, Canada) to tax gains derived from immovable property situated in that state. This exception is reinforced by art. 13(4), which provides that the source state also preserves its right to tax capital gains arising from the disposition of shares the value of which is derived principally from immovable property situated in the source state. Article 13(4) is effectively an anti-avoidance rule: it prevents companies from avoiding source state taxation and thereby contradicting art. 13(1) simply by using a corporate vehicle to hold and sell the immovable property (J. Li and F. Avella, “Article 13: Capital Gains”, in Global Tax Treaty Commentaries, May 30, 2020 (online), at s. 1.1.2.5; see also T.C.C. reasons, at para. 41).

[137]                     Finally, there is a carve-out to the exception set out by art. 13(4), for property in which the company carries on business. Capital gains arising from the disposition of such property will be taxable in the state of residence as such gains fall under the rule provided by art. 13(5). In this case, the Tax Court judge found that the carve-out applies to the disposition of the shares in Alta Canada, which can be taxed by the state of residence (Luxembourg) and not by Canada.

[138]                     Although the text of art. 13 of the Treaty sets out a clear distributive scheme to allocate taxing rights, it sheds little light on the rationale underlying this allocation. Canada and Luxembourg did not allocate taxing rights at random; there is a logic underlying the distributive scheme set out in art. 13 of the Treaty. It is this rationale that we must identify in order to give effect to the GAAR.

(b)          Context: Specific Anti-Avoidance Provisions

[139]                     Alta Luxembourg argues that the Crown’s submitted rationale, specifically what it calls a “substantial economic connection test”, has the effect of changing the bargain struck between Canada and Luxembourg. It contends that the Crown is seeking to add the unstated condition that a taxpayer must have substantial economic ties to the contracting state of residence in order to qualify as a “resident” under the Treaty. Canada could have negotiated the inclusion of a provision that could have achieved a similar effect but elected not to. Specifically, Alta Luxembourg submits that, in negotiating with Luxembourg, Canada chose not to include one — or many — of the specific anti-avoidance rules discussed in the 1998 OECD “Commentary on Article 1”, in Model Tax Convention on Income and Capital: Condensed Version. Therefore, Alta Luxembourg says, the Minister should not be allowed to rely on the GAAR to deny treaty benefits to an entity meeting the residency requirements outlined in the Treaty.

[140]                     This argument must fail as it ignores the very justifications behind the legislative choice to introduce a general anti-avoidance provision in a tax system. While there are specific rules that may have permitted the denial of the tax benefits conferred here had they been included in the Treaty, it does not follow that the GAAR cannot apply to deny the tax benefit.

[141]                     Since not every tax avoidance strategy can be foreseen, the effectiveness of specific provisions is limited to Parliament’s ability to anticipate such schemes. For this reason, those provisions cannot completely thwart inappropriate tax avoidance (Canada, Report of the Royal Commission on Taxation (1966), vol. 3, at p. 554). Before the inception of the GAAR, the increased use of complex tax avoidance schemes and matching responses to curb them left taxpayers with the expectation that purely tax-motivated transactions that were inconsistent with the rationale of the provisions relied upon were acceptable as long as they are not targeted by specific legislation (Dodge, at p. 4). Additionally, because legislation is generally prospective, specific anti-avoidance rules permit the early participants in avoidance schemes — those who can afford “the most astute tax advisers” — to enjoy the benefits of those schemes (p. 9). This may result in the loss of significant tax revenues.

[142]                     The GAAR was enacted as a much needed modernization of Canada’s tools against tax avoidance. It was meant to respond appropriately to the proliferation of such complex arrangements and to reduce the burden of having to perpetually address abusive schemes with specific legislation matching each newly marketed “purely tax-motivated scheme” (Dodge, at pp. 4 and 8). The Explanatory Notes issued with the GAAR stated that it was to apply “as a provision of last resort after the application of the other provisions of the Act, including specific anti-avoidance measures” (Department of Finance (1988), at p. 461 (emphasis added)).

[143]                     Here, Alta Luxembourg’s position essentially amounts to an argument based on the “implied exclusion” rule: if the contracting parties had meant to include an anti-avoidance provision within the Treaty, they would have done so expressly. This very reasoning was unanimously rejected by this Court in Copthorne.

[144]                     In that case, a parent and a subsidiary corporations strategically became sister corporations to conduct a horizontal, rather than vertical, amalgamation. This had the effect of increasing the paid-up capital of the shares of the amalgamated corporations compared to the paid-up capital that would have been otherwise available. The taxpayer argued that because its transactions were not caught by the detailed provisions seeking to prevent taxpayers from inappropriately preserving paid-up capital, no abusive tax avoidance could have resulted. Parliament, the argument went, would have expressly referred to such transactions if it had meant for them to be included under the provisions. This Court rejected this argument.

[145]                     Rothstein J. stated that the implied exclusion argument, while relevant in an exercise of statutory interpretation, does not align with “the nature of a GAAR analysis” and “is misplaced where it relies exclusively on the text of the [paid-up capital] provisions without regard to their underlying rationale” (paras. 109 and 111). He also warned that “[i]f such an approach were accepted, it would be a full response in all GAAR cases, because the actions of a taxpayer will always be permitted by the text of the Act” (para. 111).

[146]                     As a result, although the provisions at issue did not impose express limits on the preservation of paid-up capital on a horizontal amalgamation, the GAAR applied to deny the tax benefit. This conclusion was justified because the series of transactions allowed for the preservation of the paid-up capital upon a “horizontal” amalgamation that allowed the taxpayer to obtain tax benefits that would not have been available in a vertical amalgamation, a result contrary to the underlying rationale of the relevant provision (Copthorne, at para. 126).

[147]                     Moreover, relying on Prévost Car Inc. v. Canada, 2009 FCA 57, [2010] 2 F.C.R. 65, to suggest that Canada could have relied on the notion of beneficial ownership to prevent the tax avoidance that took place here is misplaced. Not only was Canada entitled to rely on the GAAR rather than negotiate a specific anti-avoidance rule in the Treaty, it is important to note that the provision at issue in that case was art. 10, not art. 13, dealing not with capital gains but with dividends.

[148]                     Indeed, beneficial ownership is utterly foreign to art. 13. This notion is relevant to, as noted, art. 10, as well as arts. 11 and 12 of the 2003 OECD “Model Convention with Respect to Taxes on Income and on Capital”, in Model Tax Convention on Income and on Capital: Condensed Version, dealing respectively with the payment of dividends, interest and royalties. In a paper on whether the concept of beneficial ownership should be extended to arts. 13 and 21 of the 2001 United Nations Model Double Taxation Convention between Developed and Developing Countries, U.N. Doc. ST/ESA/PAD/SER.E/213 (identical in structure to the OECD “Model Convention”), Professor P. Baker explained, in 2008, that “the inclusion of a beneficial ownership limitation in the capital gains article of specific, bi-lateral conventions is not part of the current treaty practice of any state” and that “[t]here is also something of a conceptual gulf between applying the beneficial ownership concept to items of income — such as dividends, interest and royalties — and applying it to a capital gain” (The United Nations Model Double Taxation Convention Between Developed and Developing Countries: Possible Extension of the Beneficial Ownership Concept, U.N. Doc. E/C.18/2008/CRP.2/Add.1, Ann., October 17, 2008, at paras. 54-55 (emphasis added)). While Professor Baker did state that the “theoretical differences between a flow of income — as in the case of dividends, interest and royalties — and the effective beneficiary of a capital gain, are not insurmountable”, he then went on to state that “[i]t [was] also worth bearing in mind that there is something of a conflict in legal concepts in recognising that the owner of an asset who disposes of that asset may not be the beneficial owner of the capital gain arising from the asset” (para. 58 (emphasis added)). Further, in a subsequent report, the United Nations’ committee of experts overseeing the subcommittee who had commissioned Professor Baker’s paper concluded that “[t]here was ultimately only limited support for inserting beneficial ownership in article 13” (Committee of Experts on International Cooperation in Tax Matters, Report on the fourth session (20-24 October 2008), U.N. Doc. E/2008/45 (Supp.), 2008, at para. 46).

[149]                     But, of more relevance to the present debate, the examples of tax treaty abuse through the use of art. 13 reviewed by Professor Baker involved, as here, transfers of residence before the disposal of an asset. He explained that the introduction of “[a] beneficial ownership limitation would not necessarily counter [these] transactions” (para. 53). For all these reasons, beneficial ownership would not be an appropriate anti-avoidance tool to curb tax treaty exploitation involving art. 13 dealing with the realization of capital gains.

[150]                     In sum, by Alta Luxembourg’s logic, if the lack of a specific tax-avoidance provision in the Treaty is fatal to the Crown’s position, not only would the GAAR be prevented from applying to tax treaties, it would also be prevented from applying in cases of abuse of the Act that are not specifically addressed. In our view, no inference can be drawn from the absence of an anti-avoidance rule designed to counter the particular scheme employed in this case. To do so would be inconsistent with of the role of the GAAR relative to other anti-avoidance tools. It would also be patently contrary to this Court’s repeated holding that the GAAR is a provision of last resort, and can only find application where no other provisions of the Act do (Canada Trustco, at para. 21; Copthorne, at para. 66). Thus, the absence of specific anti-avoidance rules in the relevant provisions of the Treaty sheds little light on their underlying rationale.

(c)           The Purpose of the Relevant Provisions: Economic Allegiance Underlies the Distributive Scheme

[151]                     The Treaty, like other bilateral tax treaties, allocates the right to tax between contracting states so as to avoid double taxation (E. C. C. M. Kemmeren, “Legal and Economic Principles Support an Origin and Import Neutrality-Based over a Residence and Export Neutrality-Based Tax Treaty Policy”, in M. Lang et al., eds., Tax Treaties: Building Bridges between Law and Economics (2010), 237, at pp. 239 and 243; K. Vogel, “Double Tax Treaties and Their Interpretation” (1986), 4 Int’l Tax & Bus. Law. 1, at pp. 8 and 22; Crown Forest Industries Ltd. v. Canada, [1995] 2 S.C.R. 802, at para. 46). The allocation of taxing powers follows the theory of “economic allegiance”, under which an economic connection between the state and the taxpayer serves as the basis of taxation (Li and Avella, at s. 1.1.2.1, referring to League of Nations, Economic and Financial Commission, Report on Double Taxation Submitted to the Financial Committee (1923)). This means that taxes should be paid on income where it has the strongest “economic interests” or ties, either in the state of residence or the source state.

[152]                     The theory of economic allegiance explains why arts. 1 and 4 provide that beneficiaries of the Treaty are the residents of either state. Contracting states extend the benefits of the tax treaties only to their residents (that is, persons liable to pay tax in their state) because residence is an appropriate criterion to ensure that the treaty will “cover only persons who had an economic allegiance to one or both of the contracting states” (P. J. Hattingh, “Article 1 of the OECD Model: Historical Background and the Issues Surrounding It” (2003), 57 Bull. Int’l Fisc. Doc. 215, at p. 221).

[153]                     Following the theory of economic allegiance, active income is generally taxable in the source country (where the income is earned) while passive income is taxed primarily in the residence country (J. Li and A. Cockfield, with J. S. Wilkie, International Taxation in Canada: Principles and Practices (4th ed. 2018), at pp. 46-47).

[154]                     While capital gains “do not fit nicely in either ‘active income’ or ‘passive income’” (J. Li, A. Cockfield and J. S. Wilkie, International Taxation in Canada: Principles and Practices (2nd ed. 2011), at p. 44), nonetheless the distributive scheme of art. 13 reflects, as well, the theory of economic allegiance (Li and Avella, at s. 1.1.2.1). Depending on where the economic connection is stronger, the right to tax the capital gain will be allocated to either the source state or the residence state. As a general rule, residence is taken to indicate the state to which economic ties are closest. The state of residence is typically where “capital is accumulated and consumed” (ibid.), so it has the greater claim to tax. As Professor Krishna explains, “[t]he obligation to pay tax based on residence derives from the principle that persons who benefit from their economic and social affiliation with a country have an obligation to contribute to its public finances” (Fundamentals of Canadian Income Tax (2nd ed. 2019), vol. 1, at p. 103).

[155]                     Article 13(5) of the Treaty reflects this principle: with certain exceptions, capital gains are taxable only in the state in which the taxpayer is a resident, because residence connotes this close connection.

[156]                     Gains deriving from immovable property form part of such an exception (see, generally, art. 13(1) and (4)). In relation to immovable property, the capital gain is generally connected predominantly to the situs of the property, which justifies allocating taxing rights to the source state. The rationale behind source taxation is that “a person who receives income from a person or property situated in a state has such a close relation with the state where the person or property is physically located that an obligation to support that state is justified on grounds of the relationship” (Kemmeren, at p. 262 (emphasis in original); see also A. Christians and N. Benoît-Guay, “Status and Structure of Tax Treaties”, in J.-P. Vidal, ed., Introduction to International Tax in Canada (9th ed. 2021), at p. 4/14 to 4/15). This principle explains why art. 13(1) and (4) allocates to the source state the right to tax gains derived directly or indirectly from immovable property situated in that state (Li and Avella, at ss. 1.1.2.1 and 1.1.2.5; OECD, “Commentary on Article 13”, in Model Tax Convention on Income and on Capital: Condensed Version (2017) (“2017 Commentary”), at para. 4; S. Simontacchi, Taxation of Capital Gains under the OECD Model Convention: With special regard to Immovable Property (2007), vol. 29, at p. 201).

[157]                     Finally, the carve-out of art. 13(4) allocates the taxing right to the state of residence for its residents’ capital gains from immovable property when it is driven by business activity, to reflect what Canada and Luxembourg considered to be closer economic ties with the residence state. “The rationale underlying this carve-out is that where a non-resident actively invests in immovable property situated in the source country, tax should be levied in the residence country” (Christians and Benoît-Guay, at p. 4/15 (emphasis added)). When business is carried on in an immovable property, the business activity — rather than the immovable property itself — drives value. Accordingly, the economic justification for source state taxation is weaker.

[158]                     It is worth recalling that art. 13(4) is intended to prevent taxpayers from circumventing source state taxation (art. 13(1)) by interposing a company to alienate immovable property. This justification for source state taxation is weaker when the company carries on a business in the immovable property because “the assumption that a corporate veil has been interposed between the shareholder and the immovable property is questionable, and equating the alienation of shares in the company to the alienation of the underlying immovable property does not seem tenable” (S. Simontacchi, “Immovable Property Companies as Defined in Article 13(4) of the OECD Model” (2006), 60 Bull. Int’l Tax’n 29, at p. 31).

[159]                     The carve-out is not in the 2017 OECD “Model Convention”, but it is contemplated by the OECD’s commentaries as an option for contracting parties who want to allocate taxing rights to the residence state for immovable properties such as mines and hotels (2017 Commentary, at paras. 28.5-28.7). As a result, we agree with the Tax Court that one of the purposes of the carve-out is “to encourage investments by Luxembourg residents in Immovable Property acquired to be used in a company’s business” (para. 43).

[160]                     Relying on Fundy Settlement v. Canada, 2010 FCA 309, [2012] 2 F.C.R. 374, the Federal Court of Appeal rejected the theory of “economic allegiance” as the rationale underlying art. 13(4) of the Treaty because “[t]here is no distinction in the Luxembourg Convention between residents with strong economic or commercial ties and those with weak or no commercial or economic ties” (R. v. MIL (Investments) S.A., 2007 FCA 236, [2007] 4 C.T.C. 235 (“MIL (FCA)”), at para. 6).

[161]                     In our view, the reasoning adopted in Fundy Settlement, MIL (TCC) and MIL (FCA) departs from the proper approach under the GAAR, which focuses on the “why” — the rationale — behind the text of the relevant provisions; these cases should not be followed. If relying on the text of a treaty provision is sufficient to avoid abusing a treaty, then the GAAR is effectively rendered powerless to address abusive transactions structured around tax treaties, notwithstanding the retroactive 2005 amendments to s. 245 of the Act to ensure that the GAAR applied to tax treaties. As Professor G. T. Loomer explains, this analysis “seems to ensure that any avoidance strategy involving a Canadian tax treaty, no matter how blatant, will be immune to the GAAR provided the intermediate entity is formally established in the particular state and meets the technical requirements of the particular treaty provisions relied upon” (“Tax Treaty Abuse: Is Canada Responding Effectively?”, in Oxford University Centre for Business Taxation, Working Paper No. 09/05 (revised March 2009), at s. 4.3.3). We would also note that, when this Court disposed of the appeal in Fundy Settlement on other grounds, it specified that it “should not be understood as endorsing the reasons of the Federal Court of Appeal” on the GAAR (Fundy Settlement v. Canada, 2012 SCC 14, [2012] 1 S.C.R. 520, at para. 19).

[162]                     Alta Luxembourg similarly argues that this Court held, in Crown Forest, that residence under bilateral tax conventions is established where a taxpayer is subject to full tax liability in the contracting state. In its view, the Crown is seeking to add the unstated condition that a taxpayer must have substantial economic ties to the contracting state of residence in order to qualify as a “resident” under the Treaty.

[163]                     We disagree. This argument ignores the fact that Crown Forest did not deal with the application of the GAAR. The Court in that case had to interpret the term “resident” under the Canada-United States Tax Convention Act, 1984. By contrast, where, as here, the GAAR is invoked to deny a tax benefit that would otherwise be available based on the text of the provisions, one must look beyond the text in order to determine whether conferring such benefit frustrates or exploits the underlying rationale of the provisions. This does not mean that Alta Luxembourg must meet an “unstated condition” in addition to qualifying as a “resident” under arts. 1 and 4 of the Treaty. Rather, the Crown raises the need for economic ties to Luxembourg in relation to the object, spirit or purpose underpinning the relevant provisions. For the Crown, granting the tax benefit to Alta Luxembourg in the circumstances where it fails to have any real, substantial economic connection to the state of residence would frustrate the underlying rationale of the relevant provisions. This accords with the proper methodology under the GAAR.

[164]                     In sum, the distributive scheme allocates taxing rights based on the strength of the economic connection between the income and the contracting states. The carve-out of art. 13(4) follows the same logic: when a taxpayer carries on a business in an immovable property, it derives benefits more from the residence state and that state’s commercial legal framework than from the source state’s infrastructure. This connection between the income and the residence state forms the economic basis for residence state taxation.

(d)          Conclusion on the Rationale for the Relevant Provisions

[165]                     We conclude that the rationale for the relevant provisions of the Treaty is to allocate taxation rights to the contracting state which has the strongest economic claim to the relevant income. The carve-out assigns the right to tax capital gains arising from the disposition of immovable property in which business is carried on to the resident state. The rationale behind the carve-out is to encourage investment; it reflects the fact that the business activity, rather than the immovable property itself, drives the value of the property. Accordingly, the economic justification for source state taxation — an exception to the general principle of taxation in the state of residence — is weaker. The state of residence has the stronger economic claim to tax the income.

(2)          Was There an Abuse of the Provisions of the Treaty?

[166]                     The courts below erred in law by failing to identify the object, spirit or purpose of the Treaty’s relevant provisions and, instead, merely restated art. 13(5). The courts failed to focus on the reason for the article. Consequently, we are called on to conduct the abuse analysis afresh and to determine whether the transaction falls within or frustrates the rationale of the relevant provisions of the Treaty (Canada v. 594710 British Columbia Ltd., 2018 FCA 166, [2019] 5 C.T.C. 1, at para. 64).

[167]                      In our view, Alta Luxembourg’s presence in Luxembourg is not genuine — it is mere gossamer. As such allowing it to enjoy tax benefits conferred on the basis of the stronger economic connection between the taxpayer’s income and Luxembourg frustrates the object, spirit or purpose of the provisions of the Treaty.

[168]                     Alta Luxembourg’s lack of substantial economic connection is plain from a review of the record. Alta Luxembourg was created and controlled by Alta US and its American and other foreign shareholders whose jurisdiction did not have a tax treaty with a similar exemption as the Treaty. The transaction was structured to ensure no gain would be realized in Luxembourg and that all proceeds of the sale would flow to the shareholders. As a senior managing director of the private equity firm behind the restructuring said, the objective was not to establish a genuine presence in Luxembourg, but rather to ensure that “everything will . . . flow through vehicles and no Canadian taxes will be paid” (A.R., vol. IV, at p. 106). After the transaction, Alta Luxembourg did not carry business in Luxembourg.

[169]                     Alta Luxembourg’s presence in Luxembourg was manufactured out of whole cloth. The shareholders and investors of Alta US retained “domiciliation services” to provide a corporate presence in Luxembourg and benefit from the Treaty. Email exchanges preceding the transaction make it clear that the service provider’s role was solely to ensure that Alta Luxembourg had the minimum “substance requirements” so the shareholders and investors of Alta US could “avail [themselves of] the benefits under the Canada/Luxembourg tax treaty (i.e., to alleviate Canadian tax leakage)” (A.R., vol. IV, at pp. 138-39). Calls for Alta Luxembourg would be routed to a switchboard and answered by an employee from the service provider. In addition to providing a corporate presence to Alta Luxembourg, the service provider also held board meetings for Alta Luxembourg. The members of the board of Alta Luxembourg were composed of two classes of managers: Class A managers, and Class B managers. All Class B managers were provided by the service provider and never met with Class A managers. In essence, Alta Luxembourg only existed through the service provider. It was an empty shell.

[170]                     Canada and Luxembourg agreed to the carve-out for the purpose of reflecting the closer economic connection between Luxembourg and its residents who carry on business in an immovable property situated in Canada. Yet the avoidance transactions were structured around a shell company with no real ties to Luxembourg, which was created solely to obtain a tax benefit under the Treaty. Therefore, the transactions led to an outcome that defeats the rationale underlying the allocation of taxing rights to Luxembourg over a class of capital gains. Indeed, this is precisely the type of transaction that the authors Li and Avella qualify as abusive uses of tax treaties: “. . . using interposed companies in an intermediary jurisdiction to take advantage of that jurisdiction’s treaty network . . .” in relation to capital gains (s. 2.1.1.3).

[171]                     Canada has deliberately decided not to extend the benefit of the treaty exemption found in the Treaty to the residents of the jurisdictions of the shareholders of Alta US. With regards to the provisions at issue, the common intention of Canada and Luxembourg could not have been to provide an avenue for residents of third-party states to indirectly obtain tax benefits they could not obtain directly absent any real economic connection with Luxembourg. This is “patently contrary to the basis on which” states cede their “jurisdiction to tax as the source country” (Crown Forest, at para. 52). Our colleague’s reasons assume that the federal government deliberately set out, in the exercise of its treaty making authority, to create the conditions for unlimited tax avoidance by means of schemes such as that in which Alta Luxembourg was used. To state such a proposition is to expose its absurdity, yet our colleague seeks to legitimize such blatantly abusive tax avoidance based on her view that Canada should have negotiated different treaty terms. The focus on what else could hypothetically have been agreed to is misplaced. It involves ex ante speculation about how the treaty parties ought to have proceeded based on alternatives said to have been available to them. However, such an argument gives primacy to what is not there. We are of the view that Parliament was entitled to rely on the GAAR to address abusive uses of the Treaty rather than negotiate the inclusion of a specific rule. The focus should be on what was actually agreed upon and whether the underlying rationale of the relevant provisions was frustrated by the avoidance transactions undertaken. In the give and take of treaty negotiation, Canada certainly did not give up the GAAR.

[172]                     The Auditor General raised concerns with precisely these types of schemes 20 years ago and recommended that Parliament take legislative action to protect Canada’s tax base. Parliament addressed this “mischief” with the enactment of s. 4.1 ITCIA to ensure the Minister could rely on the GAAR to deny tax benefits obtained through the use of schemes found to be abusive under the GAAR analysis. To borrow the words of Li and Cockfield, to “ascertain whether Parliament . . . intended to prohibit a particular tax-avoidance transaction”, it is wise to determine what is the “mischief” it sought to remove (p. 380).

[173]                     In addition, the avoidance transactions run contrary to the objective of encouraging trade and investment which underlies the carve-out. They were not an acquisition or an investment, but rather a liquidation of an existing investment in Canada without paying tax on the capital gain. Thus, the transactions are disconnected from the economic objectives underlying the bargain.

[174]                     Further, although “insufficient by itself to establish abusive tax avoidance”, our jurisprudence is clear that the absence of a bona fide non-tax purpose “may form part of the factual context that the courts may consider in the analysis of abusive tax avoidance allegations under s. 245(4)” (Canada Trustco, at para. 58; see also Lipson, at para. 38). The absence of bona fide non-tax purpose is relevant when the rationale of the provisions at issue means “that a particular tax benefit may apply only to transactions with a certain economic, commercial, family or other non-tax purpose” (Canada Trustco, at para. 58).

[175]                     Here, the Minister is seeking to apply the GAAR in the context of a purely tax-motivated corporate restructuring resulting in a shift of residence to Luxembourg, a reorganization for which there was no bona fide non-tax purpose. Given that the rationale of the relevant provisions is to reflect the economic ties with Luxembourg, the state of residence, this utter lack of any non-tax purpose is a relevant underlying factor that can be considered in the analysis under s. 245(4) of the Act.

[176]                     While the Crown has conceded that Alta Luxembourg was a resident of Luxembourg under art. 4 of the Treaty, contrary to the Federal Court of Appeal, we do not consider the residency requirement to be determinative in the context of a GAAR analysis. Again, the GAAR framework requires courts to give effect to the rationale underlying the text of the relevant provisions. The reasoning of the Federal Court of Appeal would effectively amount to holding that the GAAR does not apply to tax treaties when the benefit is to be exempt from taxation in Canada, i.e. when Canada has ceded its right to its treaty partner. Since being exempt from taxation in Canada is the main tax benefit provided by tax treaties, the GAAR, a general rule of last resort, would be rendered nugatory as a rule in relation to abuses of tax treaties. Technical compliance with a tax treaty in a way that frustrates the underlying rationale of the provisions relied upon by the taxpayer is precisely what triggers the GAAR. Alta Luxembourg complied with the words of the relevant provisions of the Treaty, but not their rationale. It is that to which the GAAR requires courts to give effect.

[177]                     We acknowledge that finding that a transaction structured to claim tax benefits from a treaty can be abusive when a resident lacks economic connections to the state of residence may produce more uncertainty than mechanically applying the words of the Treaty. However, Parliament struck the balance it considered proper between certainty and fairness to the tax system as a whole. The facts of this case are a patent example of a sophisticated taxpayer effecting a restructuring on the basis of professional tax advice to avoid Canadian tax. In such cases, the principle of fairness ought not to be ignored. As for the degree of uncertainty introduced by the GAAR, it is counterbalanced by the Crown’s burden to show that the avoidance transactions frustrate the object, spirit or purpose of the provisions relied on by the taxpayer and by the fact that any doubt under the GAAR analysis is to be resolved in favour of the taxpayer (Canada Trustco, at para. 69). In this case, the abuse is clear. The evidence demonstrates that Alta Luxembourg had no genuine economic connections with Luxembourg as it was a mere conduit interposed in Luxembourg for residents of third-party states to avail themselves of a tax exemption under the Treaty. We agree with our colleague that the lack of any non-tax purpose, although relevant, does not on its own lead to the determination of abuse in this case. Rather, it is this lack of any genuine economic connection to Luxembourg that frustrates the rationale of the relevant provisions of the Treaty. The Crown has discharged its burden.

[178]                   Finally, we wish to comment on an observation made at paras. 84-85 of the Tax Court decision. The Tax Court explained that a party negotiating a tax treaty is presumed to be aware of the other state’s tax system, such that Canada, as a state which taxes capital gains, should have seen to the prevention of a double exemption of this type of income. To be clear, the Crown’s position, as we understand it, is not that the GAAR is invoked because the capital gain did not go on to be taxed in Luxembourg. Rather, the Crown is seeking the application of the GAAR because in its view, Alta Luxembourg had no economic connection to Luxembourg and therefore the underlying rationale of arts. 1, 4 and 13(4) that it identified is frustrated.

[179]                     In sum, we are of the view that Alta Luxembourg’s patent lack of economic connection to Luxembourg frustrates the underlying rationale of the relevant provisions of the Treaty to allocate the taxation rights to the contracting state with the strongest economic claim. The transactions are the product of a contrivance which was not reasonably contemplated within the object, spirit or purpose of art. 13(4). As a result, the avoidance transactions frustrate the rationale of art. 13(4) and constitute abuse under the GAAR. Accordingly, the tax benefit conferred by the relevant provisions is denied.

(3)          A Word on Treaty Shopping

[180]                     We pause to say a word on whether treaty shopping constitutes an improper use of tax treaties, as this was extensively discussed by the parties and in the courts below. The OECD Glossary of Tax Terms (online) defines “treaty shopping” as follows:

An analysis of tax treaty provisions to structure an international transaction or operation so as to take advantage of a particular tax treaty. The term is normally applied to a situation where a person not resident of either the treaty countries establishes an entity in one of the treaty countries in order to obtain treaty benefits.

 

 

(See also Crown Forest, at para. 52.)

[181]                     On this issue, we agree with the Crown that the Treaty is not intended to provide avenues for residents of third-party states to benefit from their treaty partner’s tax regime in the absence of genuine ties to that state. As this Court held in Crown Forest, there is no reason to assume that Canada would enter into tax treaties to “ced[e] its taxing authority to a jurisdiction that is a stranger to the Convention” (para. 49).

[182]                     Treaty shopping connotes “a premeditated effort to take advantage of the international tax treaty network, and careful selection of the most favorable treaty for a specific purpose” (H. D. Rosenbloom, “Tax Treaty Abuse: Policies and Issues” (1983), 15 Law & Pol’y Int’l Bus. 763, at p. 766, cited in D. G. Duff, “Tax Treaty Abuse and the Principal Purpose Test — Part 1” (2018), 66 Can. Tax J. 619, at pp. 623-24). Treaty shopping typically involves the practice of non-residents establishing a minimal presence or economic activity in a country in order to benefit from the jurisdiction’s treaty network with other countries (V. Krishna, “Using Beneficial Ownership to Prevent Treaty Shopping” (2009), 56 Tax Notes Int’l 537, at p. 540).

[183]                     Treaty shopping has notably been criticized on the grounds that it produces outcomes contrary to the intention of the contracting states, such as reduced taxation and non-taxation, which can lead to an unintended erosion of national tax bases (Duff (2018), at pp. 624-25; see also B. J. Arnold and J. R. Wilson, “Aggressive International Tax Planning by Multinational Corporations: The Canadian Context and Possible Responses”, University of Calgary School of Public Policy, vol. 7, Research Paper No. 29 (2014), at p. 56). Indeed, treaty shopping can indirectly and unwittingly provide tax benefits for residents of third-party states, contrary to Canada’s intention to extend such benefits solely to the residents of the treaty partner (Department of Finance, Consultation Paper on Treaty Shopping — The Problem and Possible Solutions, August 12, 2013 (online)). As a result, treaty shopping can upset the “balance of sacrifices” that underlies a tax treaty and reduce the incentive to conclude a tax treaty. Why would a state negotiate a tax treaty with Canada and make the concessions such a process entails if its residents can access Canada’s tax treaty network through treaty shopping arrangements (Duff (2018), at p. 624; OECD, Action 6 Prevention of tax treaty abuse (online); see also J. Li, “Beneficial Ownership in Tax Treaties: Judicial Interpretation and the Case for Clarity”, in P. Baker and C. Bobbett, eds., Tax Polymath: A life in international taxation (2011), 187, at p. 191)?

[184]                     In Crown Forest, this Court unanimously stated that treaty shopping is “highly undesirable” and “patently contrary to the basis on which Canada ceded its jurisdiction to tax as the source country” (para. 52). It added that such behaviour “is not to be encouraged or promoted by judicial interpretation of existing agreements” (para. 49).

[185]                     Nevertheless, we agree with Alta Luxembourg that treaty shopping is not inherently abusive. There is nothing necessarily improper about minimizing tax liability by selecting a beneficial tax regime in making an investment in a foreign jurisdiction (Crown Forest, at para. 49). Certain jurisdictions may provide tax incentives to attract businesses and investment; as such, taxpayers are entitled to avail themselves of such benefits to minimize tax. Thus, merely selecting a treaty to minimize tax, on its own, is not abusive. In fact, it may be consonant with one of the main purposes of tax treaties: encouraging trade and investment.

[186]                     However, where taxing rights in a tax treaty are allocated on the basis of economic allegiance and conduit entities claim tax benefits despite the absence of any genuine economic connection with the state of residence, treaty shopping is, in our view, abusive. As Professors N. Bammens and L. De Broe explain, the use of “conduit companies” is disconnected from the objectives of bilateral tax treaties:

. . . tax treaties are concluded for reasons of an economic nature: the contracting states want to stimulate reciprocal commercial relations by preventing double taxation. The use of conduit companies and treaty shopping structures has very little to do with this economic objective. Treaty shopping thus upsets the balance and reciprocity of the tax treaty: in order to preserve a tax treaty’s inherent reciprocity, its benefits must not be extended to persons not entitled to them. [Emphasis added; footnotes omitted.]

 

(“Treaty Shopping and Avoidance of Abuse”, in Lang et al., Tax Treaties, 51, at p. 52; see also Li and Avella, at s. 2.1.1.3.)

[187]                     In such cases, as here, the avoidance transaction would be contrary to the objectives of bilateral tax treaties and frustrate the object, spirit or purpose of the specific provisions related to the allocation of taxing rights. Preventing such abuse is the purpose of the GAAR: “. . . most double tax treaties do not contain specific limitations on the ability of third-country residents to treaty shop [and instead] rely on the concept of beneficial ownership or on domestic anti-abuse legislation to safeguard against hollow conduits” (Krishna (2009), at p. 540). Similarly, C. A. Brown and J. Bogle are of the view that the GAAR is “[t]he primary tool to fight treaty shopping in Canada currently” (“Treaty Shopping and the New Multilateral Tax Agreement — Is it Business as Usual in Canada?” (2020), 43 Dal. L.J. 1, at p. 4).

[188]                     In conclusion, not all types of treaty shopping lead to abuse of a tax treaty. Only when an avoidance transaction frustrates the rationale of the relevant treaty provision will treaty shopping be abusive and the tax benefit denied. For instance, where contracting parties allocate taxing rights to the state of residence on the basis of economic allegiance, as in this case, treaty shopping will be abusive if the resident of a third-party state uses a conduit company to claim treaty benefits conferred by provisions requiring a genuine economic connection with the residence state. Therein lies the undermining of these provisions’ rationale clothed in a formalistic adherence to their text. Ignoring this is to render the GAAR empty of meaningful effect.

IV.         Disposition

[189]                     We would allow the Crown’s appeal and set aside the judgments of the Federal Court of Appeal and the Tax Court of Canada, with costs. Alta Luxembourg’s appeal under the Act from the Minister of National Revenue’s assessment for the 2013 taxation year should be dismissed.

 

                    Appeal dismissed with costs, Wagner C.J. and Rowe and Martin JJ. dissenting.

                    Solicitor for the appellant: Attorney General of Canada, Vancouver.

                    Solicitors for the respondent: Thorsteinssons, Toronto.