Supreme Court Judgments

SUPREME COURT OF CANADA
Citation: Deans Knight Income Corp. v. Canada, 2023 SCC 16			Appeal Heard: November 2, 2022 Judgment Rendered: May 26, 2023 Docket: 39869
Between:   Deans Knight Income Corporation Appellant   and   His Majesty The King Respondent   - and -   Attorney General of Ontario, Canadian Chamber of Commerce, Tax Executives Institute, Inc., and Agence du Revenu du Québec Interveners       Coram: Wagner C.J. and Karakatsanis, Côté, Brown,* Rowe, Martin, Kasirer, Jamal and O’Bonsawin JJ.
Reasons for Judgment: (paras. 1 to 141)	Rowe J. (Wagner C.J. and Karakatsanis, Martin, Kasirer, Jamal and O’Bonsawin JJ. concurring)
Dissenting Reasons: (paras. 142 to 197)	Côté J.

 

 

Note: This document is subject to editorial revision before its reproduction in final form in the Canada Supreme Court Reports.

 

 

^* Brown J. did not participate in the final disposition of the judgment.

 

 

 

 

Deans Knight Income Corporation                                                              Appellant

v.

His Majesty The King                                                                                 Respondent

and

Attorney General of Ontario,

Canadian Chamber of Commerce,

Tax Executives Institute, Inc., and

Agence du Revenu du Québec                                                                    Interveners

Indexed as: Deans Knight Income Corp. v. Canada

2023 SCC 16

File No.: 39869.

2022: November 2; 2023: May 26.

Present: Wagner C.J. and Karakatsanis, Côté, Brown,* Rowe, Martin, Kasirer, Jamal and O’Bonsawin JJ.

on appeal from the federal court of appeal

                    Taxation — Income tax — Tax avoidance — Application of general anti‑avoidance rule — Limitation on losses deductible from taxable income — Corporation lacking income sufficient to use non‑capital losses and other tax attributes from previous years to reduce corporate income tax — Corporation entering into transactions with other parties and deducting non‑capital losses from income earned in new investment venture — Deductions denied by Minister — Tax Court holding that transactions were tax avoidance but were not abusive under general anti‑avoidance rule — Court of Appeal concluding that transactions abusive — Whether general anti‑avoidance rule applicable to deny corporation’s deductions of non‑capital losses — Income Tax Act, R.S.C. 1985, c. 1 (5th Supp .), ss. 111(5), 245.

                    Section 111(1) (a) of the Income Tax Act  allows a taxpayer’s non‑capital losses to be carried back or forward to different taxation years to offset income in those years. However, s. 111(5)  restricts non‑capital loss carryovers for a corporation if control of the corporation has been acquired by a person or group of persons, unless it continues the same or similar business that incurred the losses. Prior to the transactions in issue, Deans Knight Income Corporation (“Deans Knight”), then operating under the name Forbes Medi‑Tech Inc. (“Forbes”), had approximately $90 million of unused non‑capital losses, scientific research and development tax expenditures, and investment tax credits but given that it was in financial difficulty, it did not have income which its past losses could offset. It entered into an investment agreement with a venture capital company, Matco, and a complex arrangement was devised to take advantage of the loss carryover deduction in s. 111(1) (a) without triggering the restriction in s. 111(5) . First, Forbe’s assets and liabilities were moved into a new parent company, Newco. Second, pursuant to the investment agreement, Matco purchased a debenture convertible into some of the voting shares and all of the non‑voting shares that Newco held in Forbes. While Newco was not obliged to sell its shares to Matco, it was promised that it would receive at least a guaranteed amount if it sold the shares or if such an opportunity did not present itself. Third, Matco would find a new business venture for Forbes, which would be used to raise money through an initial public offering (“IPO”). The profits from this venture could be sheltered by the tax attributes Forbes originally could not utilize. Other than when acting pursuant to the investment agreement, Newco and Forbes could not engage in a variety of activities without the consent of Matco. The arrangement went according to plan. Matco found a mutual fund management company, Deans Knight Capital Management, that agreed to use Forbes for an IPO through which it would raise money to invest in high‑yield debt instruments. Forbes’ name was changed to Deans Knight. The IPO and subsequent investment business succeeded. Accordingly, for its 2009 to 2012 tax years, Deans Knight deducted a majority of its non‑capital losses to reduce its tax liability.

                    The Minister reassessed Deans Knight and denied the deductions. Deans Knight objected to the reassessments and appealed to the Tax Court. Among other arguments, the Minister adopted the position that the general anti‑avoidance rule in s. 245  of the Income Tax Act  (“GAAR ”) applied to deny the deductions because the transactions constituted abusive tax avoidance. The Tax Court agreed the transactions were tax avoidance transactions that resulted in a tax benefit but held they were not abusive. On appeal, the Federal Court of Appeal held that the transactions were abusive and the GAAR  applied to deny the tax benefits. It set aside the judgment of the Tax Court and dismissed Deans Knight’s appeal of the reassessments.

                    Held (Côté J. dissenting): The appeal should be dismissed.

                    Per Wagner C.J. and Karakatsanis, Rowe, Martin, Kasirer, Jamal and O’Bonsawin JJ.: The transactions were abusive and therefore the GAAR  applies to deny the tax benefits. The object, spirit and purpose of s. 111(5) is to prevent corporations from being acquired by unrelated parties in order to deduct their unused losses against income from another business for the benefit of new shareholders. Through a complex series of transactions, Deans Knight underwent a fundamental transformation that achieved the outcome that Parliament sought to prevent, while narrowly circumventing the text of s. 111(5). Without triggering an “acquisition of control”, Matco gained the power of a majority voting shareholder and fundamentally changed Deans Knight’s assets, liabilities, shareholders and business. This severed the continuity that is at the heart of the object, spirit and purpose of s. 111(5). The result obtained by the transactions frustrated the rationale of s. 111(5) and therefore constituted abuse.

                    The GAAR  was a choice by Parliament to complement its specific anti‑avoidance efforts with the enactment of a general rule. While abusive tax avoidance can involve unforeseen tax strategies, it is more broadly designed to capture situations that undermine the integrity of the tax system by frustrating the object, spirit and purpose of the provisions relied on by the taxpayer. Some uncertainty is unavoidable when a general rule is adopted, but a reasonable degree of certainty is achieved by the balance struck within the GAAR  itself. A GAAR  analysis involves a structured, three‑step test, and asks whether (1) there was a tax benefit; (2) the transaction giving rise to the tax benefit was an avoidance transaction; and (3) the avoidance transaction was abusive.

                    Analyzing whether the avoidance transactions are abusive involves determining the object, spirit and purpose of the relevant provisions, and determining whether the result of the transactions frustrated that object, spirit and purpose. The object, spirit and purpose represents the legislative rationale that underlies specific or interrelated provisions of the Act. It is critical to distinguish the rationale behind a provision from the means chosen to give that rationale effect within the provision. The object, spirit and purpose of a provision must be worded as a description of its rationale. A court is not repeating the test for the provision or crafting a new, secondary test; rather, the object, spirit and purpose is a concise description of the rationale underlying the provision, such as why relief is being provided, the conduct that Parliament sought to encourage, or the result or mischief that Parliament sought to prevent.

                    The use of a provision’s text, context and purpose to determine the rationale differs from traditional statutory interpretation. Since in a GAAR  analysis, the search is for the rationale that underlies the words, considering the provision’s text, context and purpose ensures that the intrinsic and extrinsic evidence used to discern that rationale remains tied to the provision itself. Considering a provision’s text involves asking how it sheds light on what the provision was designed to achieve, since the language and structure of the provision can be evocative of Parliament’s underlying concerns. Courts must also consider the provision’s context, with a focus on the relationship between the provision alleged to have been abused and the particular scheme within which it operates. Understanding the provision’s purpose is central to the GAAR  analysis, and legislative history and extrinsic evidence provide insight into the rationale for specific provisions.

                    Once the object, spirit and purpose has been ascertained, the abuse analysis focuses on whether the result of the transactions frustrates the provision’s object, spirit and purpose. Avoidance transactions will be abusive where their result: is an outcome that the provisions relied on seek to prevent; defeats the underlying rationale of the provisions; or circumvents provisions in a manner that frustrates their object, spirit and purpose. Courts must go beyond the legal form and technical compliance of the transactions; they must compare the result of the transactions to the underlying rationale of the provision and determine whether that rationale has been frustrated. In coming to such a conclusion, the abusive nature of the transaction must be clear. However, there is no bar to applying the GAAR  in situations where the Act specifies precise conditions that must be met, as with a specific anti‑avoidance rule; even specific and carefully drafted provisions are not immune from abuse.

                    A review of s. 111(5)’s text, context and purpose reveals its underlying rationale. With respect to the text of the provision, s. 111(5) is a restriction on a taxpayer’s ability to make use of its non‑capital losses incurred in another taxation year. First, the text of s. 111(5) references “control”, which has been interpreted as referring to de jure control. The general test for de jure control is whether the controlling party enjoys, by virtue of its shareholdings, the ability to elect the majority of the board of directors. Second, control must be “acquired by a person or group of persons”. Third, s. 111(5) creates an exception that losses remain deductible if, after an acquisition of control, the corporation engages in the same or a similar business. Thus, the connection to past losses is severed only when control has been acquired and there is a break from the corporation’s past business. The text of s. 111(5) reflects a concern with denying loss carryovers when there is a lack of continuity within the corporation, as measured by both the identity of its controlling shareholders and its business activity.

                    A contextual analysis also sheds light on the rationale behind s. 111(5). First, s. 111(5) should be considered against the foundational principles of the Income Tax Act . Under the Act, every person, including a corporation, is a separate taxpayer, and it is a foundational principle that taxpayers are to be taxed on their own earnings. When there has been an acquisition of control and a corporation’s business ceases to operate, it can no longer be understood as the same taxpayer. Furthermore, s. 111(5)  delineates the boundaries of the benefit‑conferring provision, s. 111(1)(a). Section 111(1)(a) modifies the general rule that each taxpayer is taxed based on their income and losses within a single taxation year to allow a taxpayer to deduct non‑capital losses against income in a future or prior taxation year, but only the taxpayer who suffered the loss is entitled to deduct the loss. Section 111(5) ensures that this principle is given effect for corporations. While a corporation is still the same legal person after an acquisition of control, the identity of those behind the corporation has changed. Section 111(5) functions so that the tax benefits associated with those losses will not benefit a new shareholder base carrying on a new business. This restriction is consistent with other provisions in the Act which also treat a corporation as, effectively, a new taxpayer following an acquisition of control. There are reasons why Parliament chose the de jure control test as the standard to be used on an application of s. 111(5): it is a clearer benchmark than de facto control, meaning greater certainty for the majority of transactions, which are not tax‑motivated. However, the provision’s rationale is not fully captured by the de jure test; rather, the rationale of s. 111(5) is illuminated by related provisions which both extend and restrict the circumstances in which an acquisition of control has occurred, including by looking beyond the standard documentation under the de jure control test. These provisions suggest that de jure control is not a perfect reflection or complete explanation of the mischief that Parliament sought to address. 

                    It is also necessary to consider extrinsic evidence of Parliament’s purpose. The legislative history behind s. 111(5) illustrates that Parliament was concerned with addressing the trading of loss corporations, which was undermining the tax base and creating inequity among taxpayers. While the means Parliament has chosen to address these concerns have evolved over time, its rationale for including the non‑capital loss carryover restriction in the Act has been consistent. When a corporation changes hands, and the loss business ceases to operate, the corporation is effectively a new taxpayer that cannot avail itself of non‑capital losses accumulated by the old taxpayer. The business continuity exception was included to encourage the recovery of unprofitable enterprises that require new investment by new owners to strengthen the corporation’s business. Although the corporation may have changed hands, the link in continuity is preserved through a different marker and the justification for s. 111(1)(a) remains applicable. This reinforces that, at its core, s. 111(5) serves to delineate the circumstances in which the basis for the loss carryover rule in s. 111(1)(a) is non‑existent.

                    Taken together, the object, spirit and purpose of s. 111(5) is to prevent corporations from being acquired by unrelated parties in order to deduct their unused losses against income from another business for the benefit of new shareholders. Parliament sought to ensure that a lack of continuity in a corporation’s identity was accompanied by a corresponding break in its ability to carry over non‑capital losses. This is the rationale underlying the provision and properly explains why Parliament enacted s. 111(5).

                    An analysis of the transactions at issue demonstrates that their result served to frustrate the object, spirit and purpose of s. 111(5): they achieved the outcome that Parliament sought to prevent and provided Matco with the benefits of an acquisition of control, all while narrowly circumventing the application of s. 111(5). They resulted in Deans Knight’s near‑total transformation: it became a company with new assets and liabilities, new shareholders and a new business whose only link to its prior corporate life was the tax attributes. It was used as the vessel for an unrelated venture selected by Matco. Matco achieved the functional equivalent of an acquisition of control through the investment agreement, while circumventing s. 111(5), because the transactions dismembered the rights and benefits that would normally flow from being a controlling shareholder. First, it contracted for the ability to select Deans Knight’s directors. Second, the investment agreement placed severe restrictions on the powers of the board of directors which, but for a circuit‑breaker transaction that occurred in this case, would normally occur through a unanimous shareholders agreement and which would lead to an acquisition of de jure control. Third, the transactions allowed Matco to reap significant financial benefits, while depriving Newco, the majority voting shareholder on paper, of each of the core rights that it could ordinarily have exercised.

                    Any residual freedom that Deans Knight had was illusory and reinforces how the transactions frustrated the rationale of s. 111(5). Deans Knight’s acceptance of the corporate opportunity presented by Matco was a fait accompli because Deans Knight was prohibited from engaging in any activity other than studying and accepting the corporate opportunity, and because the consequences of refusing the opportunity were severe. As for Newco’s ability to sell its remaining shares to a party other than Matco or to opt not to sell at all, Deans Knight’s actions were already locked down by the investment agreement, and the benefits of share ownership were already negated by being subjected to Matco’s approval. The ability to receive the guaranteed amount without selling the remaining shares to Matco was important because in certain circumstances, Matco’s purchase of the shares might lead to an acquisition of de jure control. The complex series of transactions and the flexibility built into the investment agreement were necessary only because the contracting parties sought to achieve the very mischief that s. 111(5) was intended to prevent. Considering the circumstances as a whole, the result obtained by the transactions frustrated the rationale of s. 111(5).

                    Per Côté J. (dissenting): The appeal should be allowed and the Tax Court’s judgment restored. The avoidance transactions did not frustrate the rationale of s. 111(5), and therefore, do not amount to abuse. The GAAR  requires a careful balance between the interest of the taxpayer in minimizing his or her taxes through technically legitimate means and the legislative interest in ensuring the integrity of the income tax system. Despite Parliament’s unambiguous adoption of the de jure control test in s. 111(5)  of the Income Tax Act, the majority has opted for an ad hoc approach that expands the concept of control based on a wide array of operational factors. This approach invites the exercise of unbounded judicial discretion and will result in the loss‑trading restrictions in s. 111(5)  being applied to transactions on a circumstantial basis.

                    The majority’s approach to determining the object, spirit and purpose of s. 111(5) fails to account for the central principle that the GAAR  does not and cannot override Parliament’s specific intent regarding particular provisions of the Act. The GAAR  analysis rests on the same interpretive approach employed by the Court in all questions of statutory interpretation, and is little more than a specialized form of statutory interpretation to determine Parliament’s intent. It should not be assumed that the GAAR  plays a role in every transaction and in every context. There is agreement with the majority that there is no bar to applying the GAAR  in situations where the Act specifies precise conditions that must be met to achieve a particular result, as with a specific anti‑avoidance rule; however, a provision’s text can sometimes be conclusive and fully explain its underlying rationale. The key question is whether Parliament specifically intended to prevent or permit a certain type of transaction. Where an anti‑avoidance provision has been carefully crafted to include some situations and exclude others, it is reasonable to infer that Parliament chose to limit its scope accordingly. The GAAR  was intended to catch unforeseen tax strategies, but if Parliament drafts a specific anti‑avoidance provision in a way that keeps a highly foreseeable gap open, the gap is more likely to be intentional, and relying on it should not be considered abusive.

                    Section 111(5) is a specific anti‑avoidance rule that limits what would otherwise be permissible deductions under s. 111(1)(a), which allows taxpayers to deduct non‑capital losses for the purpose of computing taxable income for a taxation year. Upon an acquisition of control, s. 111(5) prevents a corporation from carrying over losses unless the business, carried on by the corporation subject to the change of control, is continued for profit or with a reasonable expectation of profit. It bars corporate acquisitions for the singular purpose of accessing tax attributes by restricting the use of those attributes if accessed through the exercise of control. Courts have determined that “control” for the purposes of the Act means de jure control. De jure control refers to the ownership of a sufficient number of shares to have a majority of votes in the election of the corporation’s board of directors. A corporation’s constating documents create de jure control because they restrain the ability of shareholders to exercise their voting power freely. In contrast, external agreements give rise to obligations that are contractual and not legal or constitutional in nature. Consequently, the distinction between de jure and de facto control lies in the breadth of factors that can be considered in determining who has control over the corporation.

                    The object, spirit and purpose of s. 111(5) is to restrict the use of tax attributes if accessed through an acquisition of de jure control. A textual, contextual and purposive analysis of s. 111(5) of the Act reveals that Parliament never intended courts to consider factors other than those related to share ownership in determining who has control over a corporation. The majority introduces the notion of functional equivalence, which treats the investment agreement as a constating document for the purposes of control. This ignores that constating documents and external agreements are enforced in radically different ways: an ordinary contract can never be functionally equivalent to a constating document. The GAAR  cannot be invoked to override Parliament’s clear intent, and the majority’s approach departs from Parliament’s clear articulation of a de jure control test for restricting losses under s. 111(5).

                    Whether an avoidance transaction is abusive is a fact‑intensive inquiry that raises a question of mixed fact and law. Absent an extricable error of law, the application of the law to the facts is subject to the standard of palpable and overriding error. No such error exists in the instant case. As de jure control is an essential element of the object, spirit and purpose of s. 111(5), the key question is whether Matco acquired de jure control of Deans Knight and the relationship between Matco and Deans Knight is the proper focus of the abuse analysis. The Tax Court’s decision was based on a combination of findings of fact and an interpretation of the investment agreement that is supported by the evidence. There is no reviewable error in the Tax Court’s conclusion that Matco did not acquire “effective” control of Deans Knight. At no point did Matco own or have a right to own enough shares to reach a majority shareholder position. The investment agreement is of no relevance as to whether Matco acquired de jure control. It did not give Matco control over Newco’s sale of the shares remaining after Matco converted the debenture or require that Matco present a sale opportunity for those shares. Parliament’s test for control is squarely focused on voting rights arising from ownership. The right to dividends is irrelevant. The only relevant incidence of ownership is voting power, something that the investment agreement did not take away. The Tax Court made a specific credibility finding on the point that Deans Knight remained a free actor throughout the transactions. Mischaracterization of s. 111(5) did not taint this important credibility finding. Matco did not acquire Deans Knight in any practical sense. Matco was only a facilitator of the transactions and did not use Deans Knight’s non‑capital losses for its own benefit. The existence of abusive tax avoidance is, at best, unclear and the benefit of the doubt should go to the taxpayer.

Cases Cited

By Rowe J.

                    Distinguished: Canada v. Alta Energy Luxembourg S.A.R.L., 2021 SCC 49; 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; referred to: Duha Printers (Western) Ltd. v. Canada, [1998] 1 S.C.R. 795; Mathew v. Canada, 2005 SCC 55, [2005] 2 S.C.R. 643; Stubart Investments Ltd. v. The Queen, [1984] 1 S.C.R. 536; Triad Gestco Ltd. v. Canada, 2012 FCA 258, [2014] 2 F.C.R. 199; Lipson v. Canada, 2009 SCC 1, [2009] 1 S.C.R. 3; The Gladwin Realty Corporation v. The Queen, 2020 FCA 142, [2020] 6 C.T.C. 185; Commissioners of Inland Revenue v. Duke of Westminster, [1936] A.C. 1; Craven v. White, [1989] A.C. 398; Canada v. Oxford Properties Group Inc., 2018 FCA 30, [2018] 4 F.C.R. 3; Fiducie Financière Satoma v. The Queen, 2018 FCA 74, 2018 D.T.C. 5052; Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 S.C.R. 235; Buckerfield’s Ltd. v. Minister of National Revenue, [1965] 1 Ex. C.R. 299; Silicon Graphics Ltd. v. Canada, 2002 FCA 260, [2003] 1 F.C. 447; Canada v. 594710 British Columbia Ltd., 2018 FCA 166, [2019] 5 C.T.C. 1; OSFC Holdings Ltd. v. Canada, 2001 FCA 260, [2002] 2 F.C. 288.

By Côté J. (dissenting)

                    Lipson v. Canada, 2009 SCC 1, [2009] 1 S.C.R. 3; Canada Trustco Mortgage Co. v. Canada, 2005 SCC 54, [2005] 2 S.C.R. 601; Commissioners of Inland Revenue v. Duke of Westminster, [1936] A.C. 1; Canada v. Alta Energy Luxembourg S.A.R.L., 2021 SCC 49; Copthorne Holdings Ltd. v. Canada, 2011 SCC 63, [2011] 3 S.C.R. 721; Minister of National Revenue v. Landrus, 2009 FCA 113, 392 N.R. 54; Buckerfield’s Ltd. v. Minister of National Revenue, [1965] 1 Ex. C.R. 299; Duha Printers (Western) Ltd. v. Canada, [1998] 1 S.C.R. 795; Minister of National Revenue v. Consolidated Holding Co., [1974] S.C.R. 419; Silicon Graphics Ltd. v. Canada, 2002 FCA 260, [2003] 1 F.C. 447; Lyrtech RD Inc. v. The Queen, 2014 FCA 267, 2015 D.T.C. 5054; Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 S.C.R. 235; Sattva Capital Corp. v. Creston Moly Corp., 2014 SCC 53, [2014] 2 S.C.R. 633.

Statutes and Regulations Cited

Business Corporations Act, CQLR, c. S‑31.1, ss. 2 “to control”, 112.

Business Corporations Act, R.S.O. 1990, c. B.16, ss. 1(5), 115(1).

Canada Business Corporations Act , R.S.C. 1985, c. C-44, ss. 2(3) , 24(3) , 102(1) , 115(3) .

Civil Code of Québec, arts. 310, 335, 336.

Income Tax Act, R.S.C. 1952, c. 148, s. 27(1)(e) [rep. & sub. 1958, c. 32, s. 12(1)], (5) [ad. idem, s. 12(2); rep. & sub. 1963, c. 21, s. 6(1); rep. & sub. 1970-72, c. 63, s. 1], (5a) [ad. 1963, c. 21, s. 6(2)].

Income Tax Act , R.S.C. 1985, c. 1 (5th Supp .), ss. 2 to 4, 111(1)(a), (5), (5.4), 245, (1) “tax benefit”, 248(10), 249(4), 251(5)(b), 256(5.1) [ad. 1988, c. 55, s. 192], (7), (8), 256.1 [ad. 2013, c. 40, s. 94].

Authors Cited

Arnold, Brian J., and James R. Wilson. “The General Anti-Avoidance Rule — Part 2” (1988), 36 Can. Tax J. 1123.

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

Canada. Department of Finance. Guidelines for Tax Reform in Canada. Ottawa, 1986.

Canada. Department of Finance. Modernizing and Strengthening the General Anti-Avoidance Rule: Consultation Paper, 2022 (online: https://www.canada.ca/content/dam/fin/documents/modernizing-strengthening-general-anti-avoidance-rule.pdf; archived version: https://www.scc-csc.ca/cso-dce/2023SCC-CSC16_1_eng.pdf).

Canada. Department of Finance. Tax Reform 1987: Income Tax Reform. Ottawa, 1987.

Canada. Department of Finance. The White Paper: Tax Reform 1987. Ottawa, 1987.

Canada. House of Commons. House of Commons Debates, vol. IV, 1st Sess., 26th Parl., October 16, 1963, pp. 3636-37.

Canada. House of Commons. House of Commons Debates, vol. V, 1st Sess., 26th Parl., November 1, 1963, p. 4287.

Canada. Privy Council Office. Guide to Making Federal Acts and Regulations, 2nd ed. Ottawa: Department of Justice Canada, 2001.

Canada. Report of the Royal Commission on Taxation, vol. 4, part C, Taxation of Income. Ottawa, 1966.

Canada. Senate. Debates of the Senate, 1st Sess., 24th Parl., August 21, 1958, pp. 650-51.

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

Duff, David G. “General Anti-Avoidance Rules Revisited: Reflections on Tim Edgar’s ‘Building a Better GAAR’” (2020), 68 Can. Tax J. 579.

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.

Duff, David G., et al. Canadian Income Tax Law, 7th ed. Toronto: LexisNexis, 2023.

Fiscal Statement, edited by Richard De Boo Limited. Toronto: Richard De Boo (special release April 28, 1980).

Flynn, Gordon W. “Tax Planning for Corporations with Net Capital and Noncapital Losses”, in Corporate Management Tax Conference, Current Developments in Measuring Business Income for Tax Purposes. Toronto: Canadian Tax Foundation, 1982, 208.

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

Krishna, Vern. The Fundamentals of Income Tax Law. Toronto: Carswell, 2009.

Leach, Campbell W. “Making the Most of Your Losses”, in Canadian Tax Papers No. 19, Corporate Management Conference. Toronto: Canadian Tax Foundation, 1960, 27.

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

Li, Jinyan. “The Misuse or Abuse Exception: The Role of Economic Substance”, in Brian J. Arnold, ed., The General Anti‑Avoidance Rule — Past, Present, and Future. Toronto: Canadian Tax Foundation, 2021, 295.

Lord, Guy, et autres. Les principes de l’imposition au Canada, 13^e éd. Montréal: Wilson & Lafleur, 2002.

McDonnell, Thomas E. “Legislative Anti‑Avoidance: The Interaction of the New General Rule and Representative Specific Rules”, in Report of Proceedings of the Fortieth Tax Conference. Toronto: Canadian Tax Foundation, 1989, 6:1.

McGuinness, Kevin Patrick. Canadian Business Corporations Law, 3rd ed. Toronto: LexisNexis, 2017.

McGurran, H. D. “Principles of Income Tax: 2. Taxable Income from a Business” (1958), 6 Can. Tax J. 455.

Riehl, Gordon W. Incorporation and Income Tax in Canada, 4th ed. Montréal: CCH Canadian, 1965.

Samtani, Pooja, and Justin Kutyan. “GAAR Revisited: From Instinctive Reaction to Intellectual Rigour” (2014), 62 Can. Tax J. 401.

Scace, Arthur R. A., and Douglas S. Ewens. The Income Tax Law of Canada, 5th ed. Toronto: Carswell, 1983.

Strain, William J., David A. Dodge and Victor Peters. “Tax Simplification: The Elusive Goal”, in Report of Proceedings of the Fortieth Tax Conference. Toronto: Canadian Tax Foundation, 1989, 4:1.

Sullivan, Ruth. The Construction of Statutes, 7th ed. Toronto: LexisNexis, 2022.

Taylor, Roger, and Marie-Claude Marcil. “Duha Printers Revisited: Issues Regarding Corporate Control” (2022), 70 Can. Tax J. 495.

                    APPEAL from a judgment of the Federal Court of Appeal (Stratas, Woods and Laskin JJ.A.), 2021 FCA 160, 460 D.L.R. (4th) 731, [2021] 5 C.T.C. 39, 2021 D.T.C. 5095, [2021] F.C.J. No. 825 (QL), 2021 CarswellNat 2893 (WL), setting aside a decision of Paris J., 2019 TCC 76, [2019] 4 C.T.C. 2001, 2019 D.T.C. 1059, [2019] T.C.J. No. 58 (QL), 2019 CarswellNat 1133 (WL). Appeal dismissed, Côté J. dissenting.

                    Barry R. Crump, Heather DiGregorio, Robert Martz and Jennie Han, for the appellant.

                    Michael Taylor and Perry Derksen, for the respondent.

                    Alexandra Clark, Dona Salmon and Jennifer Boyczuk, for the intervener the Attorney General of Ontario.

                    Steve Suarez, Laurie A. Goldbach and Elizabeth Egberts, for the intervener the Canadian Chamber of Commerce.

                    Al Meghji, Edward Rowe and Joanne Vandale, for the intervener the Tax Executives Institute, Inc.

                    Pierre Zemaitis and Josée Fournier, for the intervener Agence du Revenu du Québec.

                   The judgment of Wagner C.J. and Karakatsanis, Rowe, Martin, Kasirer, Jamal and O’Bonsawin JJ. was delivered by

 

                   Rowe J. —

TABLE OF CONTENTS
Paragraph
I.      Overview	1
II.    Facts	7
III.   Judicial History	26
A.    Tax Court of Canada, 2019 TCC 76, [2019] 4 C.T.C. 2001	26
B.    Federal Court of Appeal, 2021 FCA 160, 460 D.L.R. (4th) 731	34
IV.  Issues	39
V.    Analysis	40
A.    Background to the General Anti-Avoidance Rule	40
B.    The Relationship Between the GAAR, the Duke of Westminster Principle and Uncertainty	46
C.    Applying the GAAR	51
(1)      Tax Benefit	53
(2)      Avoidance Transaction	54
(3)      Abusive Tax Avoidance	56
(a)      The Object, Spirit and Purpose Reflects the Rationale of the Provision	58
(b)      The Provision’s Text, Context and Purpose Are Used to Determine Its Rationale	62
(c)      The Abuse Analysis Focuses on Whether the Result of the Transactions Frustrates the Provision’s Object, Spirit and Purpose	69
(d)      Summary	73
VI.   Application	75
A.    Which Provisions Are at Issue?	75
B.    What Is the Object, Spirit and Purpose of Section 111(5)?	78
(1)      The Text of the Provision	79
(2)      The Context of the Provision	84
(a)      Section 111(5) Should Be Considered Against the Foundational Principles of the Act	85
(b)      Section 111(5) Delineates the Boundaries of the Benefit-Conferring Provision, Section 111(1)(a)	86
(c)      Parliament’s Selection of Control Tests Differs Across the Act	91
(d)      The Control Test in Section 111(5) Is Expanded and Restricted by Other “Deeming” Provisions	96
(3)      The Purpose of the Provision	100
(4)      Conclusion on Object, Spirit and Purpose	113
C.    Was There an Abuse of Section 111(5)?	121
VII. Conclusion	141
Appendix

I.               Overview

[1]                             This tax appeal raises the issue of the application of the general anti-avoidance rule (the “GAAR ”) to transactions undertaken by the appellant, Deans Knight Income Corporation, to monetize non-capital losses and other deductions.

[2]                             Under the Income Tax Act ,   R.S.C. 1985, c. 1 (5th Supp .) (the “Act ”), a taxpayer’s tax burden is normally calculated based on the income and losses from that taxation year (s. 2). However, the Act  allows non-capital losses to be carried back 3 years or carried forward 20 years in order to offset income in those years (s. 111(1)(a)). This ability to carry over losses is limited: one such limit is that, if control of the corporation has been acquired, non-capital losses from before the acquisition cannot be carried over, unless the corporation continues the same or similar business that incurred the losses (s. 111(5)). An acquisition of control occurs where a person or group of persons acquires de jure control, which generally involves acquiring sufficient share ownership to elect a majority of the board of directors (Duha Printers (Western) Ltd. v. Canada, [1998] 1 S.C.R. 795). It is also deemed to occur when a taxpayer acquires a right to acquire such shares if the purpose is to avoid the application of the loss carryover restriction (ss. 256(8) and 251(5)(b); see Appendix).

[3]                             The appellant sought to take advantage of the loss carryover rule in s. 111(1)(a) without triggering the restriction in s. 111(5). Although the transactions in this appeal will be explained in detail in the following section, a brief summary is warranted. Prior to the transactions at issue, the appellant was a struggling Canadian corporation that had approximately $90 million of unused non-capital losses, scientific research and development tax expenditures (“SR&ED” expenditures), and investment tax credits (“ITCs”) (collectively, the “Tax Attributes”). Given that it was in financial difficulty, it did not have income which its past losses could offset. It sought to monetize their value and entered into an agreement with a venture capital company, Matco Capital Ltd. (“Matco”), in order to do so. A complex arrangement was devised involving the following key transactions. First, all of the appellant’s assets and liabilities would be moved into its newly created parent company. Second, Matco would obtain a debenture which could be converted into shares of the appellant, and the appellant’s parent company was promised that it would receive at least a guaranteed amount for the sale of its remaining shares. Third, Matco would find a new business venture for the appellant, which would be used to raise money through an initial public offering (“IPO”). The profits from this venture could be sheltered by the Tax Attributes the appellant originally could not utilize. The arrangement went according to plan. For the 2009 to 2012 tax years, the appellant deducted a majority of its Tax Attributes to reduce its tax liability. However, the Minister of National Revenue reassessed and denied these deductions.

[4]                             Before this Court, the parties accept that the appellant complied with the text of the Act . In other words, the parties agree that there was no “acquisition of control” and that, therefore, the loss carryover restriction in s. 111(5) did not apply. The central issue in this appeal is whether s. 245  of the Act , known as the general anti-avoidance rule or the GAAR , applies to deny the deductions. The GAAR  operates to deny tax benefits flowing from transactions that comply with the literal text of the Act  but nevertheless constitute abusive tax avoidance. For the GAAR to apply to a transaction, three elements found in s. 245 must be met: (1) there must be a “tax benefit”; (2) the transaction must be an “avoidance transaction”, meaning one that is not undertaken primarily for a bona fide non-tax purpose; and (3) the avoidance transaction giving rise to the tax benefit must be an “abuse” of the provisions of the Act  (or associated enactments).

[5]                             The Tax Court found that the transactions were tax avoidance transactions that resulted in a tax benefit, but concluded that they were not abusive. On appeal, the Federal Court of Appeal held that the transactions were abusive, such that the GAAR  applied to deny the tax benefits. I note that the parties and the lower courts focused on the non-capital loss deductions since the SR&ED and ITC provisions function similarly. As was the case in the Federal Court of Appeal, the only issue on appeal is whether the appellant’s series of transactions resulted in abusive tax avoidance.

[6]                             For the reasons that follow, I would dismiss the appeal. The transactions were abusive. The object, spirit and purpose of s. 111(5)  of the Act  is to prevent corporations from being acquired by unrelated parties in order to deduct their unused losses against income from another business for the benefit of new shareholders. Through a complex series of transactions, the appellant underwent a fundamental transformation that achieved the outcome that Parliament sought to prevent, while narrowly circumventing the text of s. 111(5). The result of the transactions thereby frustrated the provision’s rationale. Since the GAAR applies to deny the tax benefits, the Minister’s reassessments must be restored.

II.            Facts

[7]                             Before the transactions at issue in this appeal, the appellant carried on a drug research and nutritional food additive business under the name Forbes Medi-Tech Inc. Its shares were publicly listed on the NASDAQ and TSX. In 2007, the appellant’s business was struggling and it faced a potential delisting of its shares from the NASDAQ because the bid price for its common stock had fallen below the minimum price required by the exchange. At a meeting of the appellant’s board of directors in May 2007, then-CFO David Goold reported on a method of realizing the value of its accumulated Tax Attributes in the form of non-capital losses, SR&ED tax expenditures and ITCs. These unused Tax Attributes had accumulated to nearly $90 million by the end of 2007. Goold told the board that a reorganization of the company, followed by a takeover by another company, would allow the Tax Attributes to be monetized for between 4 and 4.5 cents on the dollar for a total of between $3.5 million and $4 million. It was unlikely that the appellant would be able to use the Tax Attributes on its own at any point; indeed, by November 2007, it had only six months of cash flow left.

[8]                             The appellant entered negotiations with Matco and signed a letter of intent in November 2007. However, the venture that Matco had wanted to offset with the appellant’s Tax Attributes fell apart and in December 2007, Matco informed the appellant that it would not be proceeding with the agreement.

[9]                             In early 2008, the appellant reorganized and restructured by way of a court-approved Plan of Arrangement. The board of directors had determined that this was the best way of maintaining compliance with the NASDAQ’s minimum bid price listing standard while facilitating a future monetization of its Tax Attributes. A new company, 0813361 B.C. Ltd. (“Newco”), was incorporated, and all outstanding common shares, options and warrants of the appellant were exchanged for common shares and warrants of Newco on an 8:1 basis. The appellant thus became a wholly owned subsidiary of Newco. Newco’s shares began to be traded on the NASDAQ in substitution for the shares of the appellant.

[10]                         On March 4, 2008, the appellant and Matco entered into a second letter of intent. On March 19, 2008, the appellant entered into an investment agreement with Newco and Matco (the “Investment Agreement”).

[11]                         Under the Investment Agreement, Newco would receive roughly $3.8 million in the following manner. First, Matco purchased a convertible debenture for $3 million, subject to adjustments (ss. 2.2 and 2.3 of the Investment Agreement, reproduced in A.R., vol. II, at p. 86). The debenture would be convertible into 35 percent of the voting shares and 100 percent of the non-voting shares that Newco held in the appellant (i.e., 79 percent of the equity shares in the appellant). Second, Matco guaranteed that Newco would be able to sell its remaining shares in the appellant for a minimum of $800,000 (the “Guaranteed Amount”), subject to adjustments (s. 5.5). The remaining shares represented a majority (65 percent) of the voting shares of the appellant. Newco was not obliged to sell its shares to Matco. If an opportunity for Newco to sell its remaining shares did not present itself during the relevant period, then Matco would still be required to pay the Guaranteed Amount.

[12]                         The appellant was to be reorganized: its assets, liabilities and the amount paid by Matco for the convertible debenture would be transferred to Newco (s. 3.2). Newco would also use its best efforts to ensure that the only three directors of the appellant would be Charles Butt (President and CEO of the appellant), Goold and a representative selected by Matco. Butt and Goold would resign following the acceptance of a corporate opportunity (s. 3.4).

[13]                         Matco would have one year to present this corporate opportunity — a business opportunity that would be suitable for the appellant to commence, involving a new business and likely a new management team. This would be the business that would generate profits against which the Tax Attributes would be deducted. Newco could accept or refuse the corporate opportunity within a short timeframe (s. 4.1), but if it refused, Matco would be relieved from paying the Guaranteed Amount (s. 5.5(d)).

[14]                         If an acquisition of control of the appellant or Newco occurred, then Matco would be relieved from paying the Guaranteed Amount, and Newco would be required to repurchase the convertible debenture from Matco and to pay an additional $1 million to Matco (s. 5.5(f)). Moreover, other than when acting pursuant to the Investment Agreement, Newco and the appellant could not engage in the following activities without the consent of Matco (s. 6.1):

-           issue any shares, options, warrants, calls, conversion privileges or rights of any kind to acquire any shares of the appellant,

-           sell, transfer, pledge, encumber or dispose of or agree to sell, transfer, pledge, encumber or dispose of any shares of, or any options, warrants, calls, conversion privileges or rights of any kind to acquire any shares of the appellant,

-           change or amend the appellant’s constating documents or by-laws,

-           split, combine or reclassify any outstanding shares of the appellant,

-           redeem or purchase any shares of the appellant,

-           reorganize, amalgamate or merge the appellant,

-           take any action or make any commitment with respect to, or in contemplation of, any complete or partial liquidation, dissolution or other winding-up of the appellant,

-           declare and/or pay dividends or reduce the capital of the appellant,

-           take any action that would or may give rise to a change of control of Newco or the appellant, other than in specific circumstances contemplated in the Investment Agreement,

-           enter into, assign, terminate or amend any contract or agreement in respect of the appellant,

-           create any encumbrance on any of the assets of the appellant,

-           in respect of the appellant, create, incur, guarantee, or assume any indebtedness for borrowed money or otherwise become liable or responsible for the obligations of any other person,

-           in respect of the appellant, make any loans, advances, or capital contributions to, or investments in, any other person,

-           change in any respect any of the accounting principles or practices used by Newco or the appellant, except for any change required by reason of a concurrent change in policy, and

-           engage in any activity other than examining and pursuing the corporate opportunity.

[15]                         Before the Investment Agreement was executed, the managing director of Matco, Alan Ross, purchased 100 shares of the appellant from Newco through his wholly owned holding company, 1250280 Alberta Ltd. One of the purposes for this was to ensure that the Investment Agreement would not constitute a unanimous shareholder agreement.

[16]                         The Investment Agreement was executed on May 9, 2008. Pursuant to the agreement, the appellant’s assets and liabilities were transferred to Newco in exchange for a promissory note, which the appellant transferred to another subsidiary of Newco; Matco subscribed for the convertible debenture in the amount of nearly $3 million, which the appellant transferred to the other subsidiary. As planned, all of the appellant’s directors resigned except Butt, and Goold and Ross were elected directors.

[17]                         Matco sought to find a business that could use the appellant’s Tax Attributes. In December 2008, Matco presented the appellant with a corporate opportunity pursuant to the Investment Agreement. Deans Knight Capital Management (“DKCM”), a mutual fund management company, was interested in investing in high-yield debt instruments, which were selling at low prices because of the 2008 financial crisis. DKCM planned to raise money for the investments through an IPO. Matco proposed that DKCM would use the appellant as the corporate vehicle for the intended IPO, rather than incorporating a new company, because the appellant’s Tax Attributes would shelter the majority of the portfolio income and capital gains.

[18]                         The appellant’s board of directors discussed the proposal, did some investigation into DKCM, and approved the proposal. In December 2008, DKCM and the appellant entered into a letter of intent. The letter specified that the appellant must have at least $95 million in deductible amounts available to be used against income earned by the corporation in Canada. To allow the appellant to be used for DKCM’s business venture, DKCM would be appointed to manage the appellant; 4 of the 5 directors of the appellant would be appointed by DKCM; and the appellant would be used in a $100 million minimum IPO, whose proceeds would be used to purchase corporate debt securities that would generate income and gains that could be sheltered by the Tax Attributes. The IPO would be priced such that the appellant’s existing common shares (which would ultimately be held by Matco following its exercise of the convertible debenture) would be attributed a net asset value of $5 million.

[19]                         In February 2009, the appellant’s name was changed to its current name, “Deans Knight Income Corporation”. DKCM’s President became a director of the appellant. In March 2009, Matco’s managing director and four nominees of DKCM were appointed as directors of the appellant, and three officers of DKCM were appointed as officers of the appellant.

[20]                         Immediately prior to the IPO, Matco converted its debenture into 35 percent of the appellant’s voting shares and 100 percent of its non-voting shares. Matco also obtained an exception to the post-IPO lock-up period to enable it to purchase the remaining shares from Newco within the time period contemplated under the Investment Agreement.

[21]                         The IPO closed on March 18, 2009. Notably, the prospectus indicated that there was a risk that the Canada Revenue Agency could “successfully challenge the amount of such tax attributes or their availability to the Company” (A.R., vol. III, at p. 30). A total of 10,036,890 shares were issued at $10 per share, for proceeds of over $100 million. At this valuation, Matco’s shares of the appellant were worth over $4 million.

[22]                         In April 2009, Matco, through a related corporation, made an offer to Newco to purchase the remaining shares at the Guaranteed Amount. Though the Guaranteed Amount was a discount to the IPO price, Newco accepted the offer because it needed money for its own operations and because its board believed the share price might decrease before the end of the post-IPO lock-up period.

[23]                         The appellant’s investment business succeeded and it paid regular dividends to its shareholders in the first four years of operation. It began to wind up operations in its fifth year, as intended.

[24]                         When filing its tax returns for 2009 to 2012 and in computing its income, the appellant claimed Tax Attributes from 2007 and earlier. The appellant deducted nearly $65 million of its Tax Attributes to reduce its tax liability from the debt-securities business.

[25]                         The Minister reassessed these taxation years to disallow the claimed losses and expenditures. The appellant objected to the reassessments and appealed to the Tax Court.

III.         Judicial History

A.           Tax Court of Canada, 2019 TCC 76, [2019] 4 C.T.C. 2001

[26]                         At the Tax Court, Paris J. was faced with two issues. The first issue was whether Matco had obtained an option to purchase the majority of the voting shares of the appellant, thereby acquiring control pursuant to ss. 256(8)  and 251(5) (b) of the Act . Paris J. concluded that Matco had not obtained such a right, and, accordingly, that there had been no “acquisition of control” triggering the application of s. 111(5). This finding was not challenged on appeal and is therefore not before this Court.

[27]                         As for the second issue, the question was whether the GAAR  could apply to deny the deduction of the Tax Attributes. This required Paris J. to determine whether there was a series of transactions that resulted in a tax benefit, whether the transactions were primarily for tax avoidance purposes and whether they resulted in an abuse of the provisions of the Act . Paris J. chose to focus on the non-capital loss provisions, since the SR&ED and ITC streaming restrictions operated in a similar manner. Applying each step of the GAAR  test, Paris J. found that there was a series of transactions that resulted in a reduction of tax. He also concluded that the primary purpose of the Investment Agreement, the restructuring of the appellant and all related transactions was to monetize the Tax Attributes. Consequently, the series of transactions could be characterized as avoidance transactions. However, Paris J. concluded that the transactions were not abusive because they did not frustrate the object, spirit and purpose of the provisions of the Act . 

[28]                         First, Paris J. considered the object, spirit and purpose of ss. 111(1)(a), 111(5) and 256(8). Regarding s. 111(1)(a), he concluded that the object, spirit and purpose was to “provide relief to taxpayers who have suffered losses, given that the government, through income tax, shares in the income of a taxpayer” (para. 99).

[29]                         As for s. 111(5), Paris J. wrote that the acquisition of control test was the “means by which Parliament has determined that a loss has notionally been transferred to an unrelated party” (para. 103). He indicated that the notion of control was central to the working of s. 111(5) and that it has long been held to mean de jure control. He also considered the history of s. 111(5) and acknowledged that Parliament introduced provisions that deem de jure control to exist or not exist in some circumstances, thereby allowing the Minister to look “beyond the share registry of the corporation to determine who in substance has control” (paras. 111 and 115). However, he noted that the de facto control test in s. 256(5.1) was not adopted. Paris J. moved to the purpose of s. 111(5) and found it to be “clear that subsection 111(5) was enacted to prevent tax loss trading. The restriction on the use of losses is subject to limited exceptions relating to the rehabilitation of the loss business and to the transfer of losses between corporations under common control” (para. 126). He recognized that the underlying rationale for the choice to deny loss carryovers is that “after the acquisition of control, the corporation can be likened to a new taxpayer because it has different shareholders” (para. 128). He also indicated that the acquisition of control test served as a reasonable marker between “situations where the corporation is a free actor in a transaction and when it is only a passive participant whose actions can be manipulated by a new person or group of persons in order to utilize the losses or Tax Attributes of the corporation for their own benefit” (para. 134).

[30]                         Based on this analysis, Paris J. concluded that the object, spirit and purpose of s. 111(5) was to “target manipulation of losses of a corporation by a new person or group of persons, through effective control over the corporation’s actions” (para. 134). He also briefly considered the object, spirit and purpose of s. 256(8) and found that it was to “prevent a taxpayer from circumventing the listed avoidance provisions by acquiring control over shares or share voting rights in order to achieve effective control of the corporation” (para. 138).

[31]                         Turning to whether the object, spirit and purpose of the provisions were frustrated by the avoidance transactions, Paris J. framed the question as “whether despite there being no actual acquisition of de jure control by Matco, Matco acquired effective control of the Appellant such that the object, spirit and purpose of subsection 111(5) and the related tax attribute streaming restrictions was circumvented” (para. 144). The judge rejected the argument that a change of management, business activity, assets and liabilities, and name were relevant to determining whether Matco gained effective control. He thereafter focused on whether Matco had effective control over the majority of the voting shares of the appellant prior to the IPO and answered in the negative. He determined that “[t]he Appellant participated freely in the transactions that resulted in the use of the Tax Attributes against the investment business income” (para. 152).

[32]                         Nor did Paris J. accept that the transactions frustrated the object, spirit and purpose of s. 256(8)  of the Act . He rejected the Crown’s submission that Matco had effective control over the remaining shares of the appellant held by Newco: Newco could have sold the shares without Matco’s consent; the Investment Agreement did not provide that only Matco could present an opportunity for Newco to sell its remaining shares; and the restrictions placed on the appellant’s activities did not amount to control over its shares.

[33]                         In light of his conclusion that the GAAR  did not apply, Paris J. allowed the appeal from the reassessments.

B.            Federal Court of Appeal, 2021 FCA 160, 460 D.L.R. (4th) 731

[34]                         Before the Federal Court of Appeal, the only issue in dispute was the third step of the GAAR  analysis, namely, whether the avoidance transactions were abusive. Woods J.A., writing for a unanimous court, allowed the Minister’s appeal and found that the transactions frustrated the object, spirit and purpose of s. 111(5)  of the Act .

[35]                         Woods J.A. agreed with Paris J.’s approach to ascertaining the object, spirit and purpose of s. 111(5). However, she replaced the term “effective control” with “actual control”, since the former term had led to confusion. Thus, she determined that the object, spirit and purpose of s. 111(5) was to “restrict the use of specified losses, including non-capital losses, if a person or group of persons has acquired actual control over the corporation’s actions, whether by way of de jure control or otherwise” (para. 72 (emphasis added)).

[36]                         Woods J.A. rejected the argument that the object, spirit and purpose of s. 111(5) was fully expressed by its text. She noted that the provision was introduced to prohibit trafficking in shares of companies with loss carryovers and that the GAAR  itself was, in part, a response to the unintended use of loss carryovers. Moreover, she cited this Court’s recognition that “the general policy of the Income Tax Act is to prohibit the transfer of losses between taxpayers, subject to specific exceptions” and that “[t]his policy is . . . to be taken into account in determining Parliament’s intent” (para. 81, citing Mathew v. Canada, 2005 SCC 55, [2005] 2 S.C.R. 643, at para. 49). Woods J.A. recognized that her formulation of the object, spirit and purpose of s. 111(5) “does include forms of de jure and de facto control” but indicated that the actual control test is different from the de facto control test (para. 83).

[37]                         Turning to whether the avoidance transactions frustrated the object, spirit and purpose of s. 111(5), Woods J.A. concluded that the terms of the Investment Agreement gave Matco actual control over the actions of the appellant, both in general and in approving the corporate opportunity. At a general level, the restrictions in the Investment Agreement resulted in control being handed to Matco. As for approving the corporate opportunity, Woods J.A. remarked that “there was no realistic chance that a Corporate Opportunity would be rejected” (para. 104) because the Guaranteed Amount would have been forfeited. Although the appellant discussed the proposal and did some investigation before approving it, it was not a free actor: Woods J.A. highlighted that it was a limited investigation to ensure that “this company was not a fly-by-night operation” (para. 108).

[38]                         Woods J.A. concluded that s. 111(5) had been abused. Accordingly, the conditions for the application of the GAAR  were met and the tax benefit should be denied. As a result, the Federal Court of Appeal allowed the appeal and set aside the judgment of the Tax Court.

IV.         Issues

[39]                         Only the third step of the GAAR  analysis is challenged before this Court. The issues can therefore be stated as follows:

(1) Did the Federal Court of Appeal err in its articulation of the object, spirit and purpose of s. 111(5)  of the Act ?

(2) Did the Federal Court of Appeal err in concluding that the avoidance transactions were abusive?

V.           Analysis

A.           Background to the General Anti-Avoidance Rule

[40]                         The present appeal is not the first time this Court has considered the GAAR . A review of its origins and role within the Act  nonetheless provides useful background.

[41]                         In 1988, Parliament enacted the GAAR  in s. 245  of the Act , partially in response to this Court’s decision in Stubart Investments Ltd. v. The Queen, [1984] 1 S.C.R. 536. In Stubart, this Court rejected a literal approach to interpreting the Act . At the same time, it also rejected an interpretation of a precursor to the GAAR  that would have required transactions to have a bona fide business purpose. Instead, it offered guidelines to limit unacceptable tax avoidance arrangements. Parliament viewed the decision in Stubart as an inadequate approach to the problem of tax avoidance (Canada Trustco Mortgage Co. v. Canada, 2005 SCC 54, [2005] 2 S.C.R. 601, at para. 14).

[42]                         Moreover, abusive tax avoidance had become a problem of significant concern for Parliament. Taxpayers, aided by expert advice, increasingly devised complex legal transactions to avoid tax in ways unintended by Parliament. Once the avoidance mechanisms relied on became evident, either from advance ruling requests or tax assessments, Parliament would react to “plug” the loopholes in the Act  to prevent future use. The problem was that increasingly convoluted rules were vulnerable, creating new loopholes to exploit. This Court described this cycle in Stubart as “the action and reaction endlessly produced by complex, specific tax measures aimed at sophisticated business practices, and the inevitable, professionally-guided and equally specialized taxpayer reaction” (p. 580; see also D. A. Dodge, “A New and More Coherent Approach to Tax Avoidance” (1988), 36 Can. Tax J. 1, at p. 4). As this “cycle of action and reaction” between creative tax planners and Parliament continued, the Act  grew in size and complexity (Department of Finance, Guidelines for Tax Reform in Canada (1986), at p. 7).

[43]                         Despite these efforts, Parliament was unable to curb the proliferation of tax avoidance schemes. Corporate tax revenues in 1985-86 “were $1.2 billion lower than the initial budgetary forecasts”, a shortfall which “was considered to be caused largely by the unexpected application of loss carryforwards” (Dodge, at p. 3; see also W. J. Strain, D. A. Dodge and V. Peters, “Tax Simplification: The Elusive Goal”, in Report of Proceedings of the Fortieth Tax Conference (1989), 4:1, at pp. 4:43 and 4:52‑53).

[44]                         The GAAR  was Parliament’s chosen mechanism to interrupt this cycle. In the 1987 White Paper on Tax Reform, the government recognized that specific anti-avoidance rules are “not always desirable” because they “make the tax system more complex; they sometimes create additional unintended loopholes, and they do not deal with transactions completed before the amendments become effective” (see Department of Finance, The White Paper: Tax Reform 1987 (1987), at p. 57; see also B. J. Arnold and J. R. Wilson, “The General Anti-Avoidance Rule — Part 2” (1988), 36 Can. Tax J. 1123, at p. 1140). A new general anti-avoidance rule was meant to overcome some of these disadvantages. The novel approach found within the GAAR  explains why, upon its enactment, Parliament was able to remove certain specific anti-avoidance rules that it felt were sufficiently addressed by a general rule (White Paper, at p. 57; Triad Gestco Ltd. v. Canada, 2012 FCA 258, [2014] 2 F.C.R. 199, at paras. 52‑53; Department of Finance, Tax Reform 1987: Income Tax Reform (1987), at p. 129; Dodge, at p. 8; Department of Finance, Modernizing and Strengthening the General Anti-Avoidance Rule: Consultation Paper, 2022 (online), at pp. 16 and 19-20; see also Arnold and Wilson, at p. 1148).

[45]                         In light of the foregoing, the GAAR  is best understood as a way to overcome the disadvantages of a system based solely on specific rules (White Paper, at p. 57; Dodge, at p. 8). The GAAR  was a choice, made by Parliament, to complement its specific anti-avoidance efforts with the enactment of a general rule. To achieve this aim, the GAAR  “draws a line between legitimate tax minimization and abusive tax avoidance” (Trustco, at para. 16). Abusive tax avoidance can involve unforeseen tax strategies (Canada v. Alta Energy Luxembourg S.A.R.L., 2021 SCC 49, at para. 80). For example, in Alta Energy, this Court treated evidence of Parliament’s knowledge and acceptance of the tax strategy at issue as a relevant consideration when ascertaining its intent. However, the GAAR  is not limited to unforeseen situations; as this Court has explained, it is designed to capture situations that undermine the integrity of the tax system by frustrating the object, spirit and purpose of the provisions relied on by the taxpayer (Lipson v. Canada, 2009 SCC 1, [2009] 1 S.C.R. 3, at para. 2; Copthorne Holdings Ltd. v. Canada, 2011 SCC 63, [2011] 3 S.C.R. 721, at paras. 71-72; see also The Gladwin Realty Corporation v. The Queen, 2020 FCA 142, [2020] 6 C.T.C. 185, at para. 85; D. G. Duff, “General Anti-Avoidance Rules Revisited: Reflections on Tim Edgar’s ‘Building a Better GAAR’” (2020), 68 Can. Tax J. 579, at p. 591).

B.            The Relationship Between the GAAR, the Duke of Westminster Principle and Uncertainty

[46]                         The GAAR  must also be understood in light of its relationship to the Duke of Westminster principle. In Commissioners of Inland Revenue v. Duke of Westminster, [1936] A.C. 1 (H.L.), Lord Tomlin recognized the foundational principle that “[e]very man is entitled if he can to order his affairs so as that the tax attaching under the appropriate Acts is less than it otherwise would be” (p. 19). The principle that taxpayers can order their affairs to minimize the amount of tax payable has been affirmed by this Court on numerous occasions (see, e.g., Stubart, at p. 552; Trustco, at para. 11; Copthorne, at para. 65).

[47]                         The Duke of Westminster principle, however, has “never been absolute” (Lipson, at para. 21) and it is open to Parliament to derogate from it. Parliament has done so through the GAAR . The GAAR  does not displace the Duke of Westminster principle for legitimate tax planning. Rather, it recognizes a difference between legitimate tax planning — which represents the vast majority of transactions and remains unaffected, consistent with the Duke of Westminster principle — and tax planning that operates to abuse the rules of the tax system — in which case the integrity of the tax system is preserved by denying the tax benefit, notwithstanding the transactions’ compliance with the provisions relied upon. Even where the purpose of a transaction is to minimize tax, taxpayers are allowed to carry it out unless it results in an abuse of the provisions of the Act  (Lipson, at para. 25). Where the transaction is shown to be abusive, the Duke of Westminster principle is “attenuated” by the GAAR  (Trustco, at para. 13).

[48]                         In establishing a general anti-avoidance rule that operated to deny tax benefits on a case-specific basis, Parliament was cognizant of the GAAR ’s implications for the level of certainty in tax planning. Parliament sought to balance “the protection of the tax base and the need for certainty for taxpayers” (Department of Finance, Explanatory Notes to Legislation Relating to Income Tax (1988), at p. 461). The GAAR  was enacted to be “a provision of last resort” to address abusive tax avoidance only and was therefore not designed to create more generalized uncertainty in tax planning (Trustco, at para. 21; Copthorne, at para. 66). Some uncertainty is unavoidable when a general rule is adopted (Dodge, at p. 21; Copthorne, at para. 123). However, a reasonable degree of certainty is achieved by the balance struck within the GAAR  itself.

[49]                         First, as Professor Jinyan Li noted, “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” (“‘Economic Substance’: Drawing the Line Between Legitimate Tax Minimization and Abusive Tax Avoidance” (2006), 54 Can. Tax J. 23, at p. 40). The GAAR  only scrutinizes transactions motivated by tax avoidance, and even a tax-motivated transaction that is consistent with the object, spirit and purpose of the provisions of the Act  is unaffected by the GAAR  (see Explanatory Notes, at p. 461). By virtue of the rigorous analysis required by s. 245 , the GAAR  only affects a small subset of transactions, largely conducted by sophisticated parties with the ability to properly evaluate the risks inherent in a GAAR  reassessment. Indeed, this is precisely what occurred in the present case: the prospectus relating to the appellant’s IPO expressly recognized the risk of a successful challenge to the use of the Tax Attributes.

[50]                         Second, a proper application of the GAAR  methodology serves to ensure reasonable certainty in tax planning (P. Samtani and J. Kutyan, “GAAR Revisited: From Instinctive Reaction to Intellectual Rigour” (2014), 62 Can. Tax J. 401, at p. 403). The GAAR  is not a tool to sanction conduct that courts find immoral (Copthorne, at para. 65; Alta Energy, at para. 48). Rather, courts must conduct an “objective, thorough and step-by-step analysis” (Copthorne, at para. 68). Within this analysis, the principles of certainty, predictability and fairness do not play an independent role; rather, they are reflected in the carefully calibrated test that Parliament crafted in s. 245 of the Act and in its interpretation by this Court. It is to this test that I now turn.

C.            Applying the GAAR

[51]                         As Rothstein J. wrote in Copthorne, “[i]t is relatively straightforward to set out the GAAR  scheme. It is much more difficult to apply it” (para. 32). This is because the GAAR  confers upon courts the “unusual duty of going behind the words of the legislation” (para. 66). While the duty imposed by the GAAR  is unusual, the analysis involves a structured, three-step test that has been the subject of thorough guidance by this Court. In order for the GAAR  to apply, the following questions must be asked (para. 33, citing Trustco, at paras. 18, 21 and 36):

1. Was there a tax benefit? . . .

2. Was the transaction giving rise to the tax benefit an avoidance transaction? . . .

3. Was the avoidance transaction giving rise to the tax benefit abusive?

[52]                         The taxpayer bears the burden of refuting the Minister’s assumption of the existence of a tax benefit (Copthorne, at para. 34; Trustco, at para. 63) and the burden of proving the existence of a bona fide non-tax purpose (Copthorne, at para. 63; Trustco, at para. 66). In contrast, at the third step, the Minister bears the burden of proving that the avoidance transaction results in an abuse (Lipson, at para. 21).

(1)          Tax Benefit

[53]                         The first step in the GAAR  analysis is to determine whether a tax benefit arises from a transaction or a series of transactions (s. 245(2)  of the Act ; Trustco, at para. 18). A tax benefit is defined as a “reduction, avoidance or deferral of tax” or “an increase in a refund of tax or other amount” paid under the Act  (s. 245(1)  of the Act ). The existence of a tax benefit may be clear on the face of the transactions; it is also appropriate to have regard to alternative arrangements that “might reasonably have been carried out but for the existence of the tax benefit” (Copthorne, at para. 35, citing D. G. Duff et al., Canadian Income Tax Law (3rd ed. 2009), at p. 187).

(2)          Avoidance Transaction

[54]                         The second step in the GAAR  analysis is to determine whether the transaction or series of transactions was made primarily for the purpose of obtaining a tax benefit (s. 245(3)  of the Act ; Trustco, at paras. 17, 21 and 66). This requirement removes the majority of transactions from the ambit of the GAAR , including those that are made for family and investment purposes (Trustco, at paras. 21 and 33). A transaction may have both tax and non-tax purposes; in such a case, the taxpayer must satisfy the court that it is reasonable to conclude that the non-tax purpose was primary (paras. 27 and 29).

[55]                         Moreover, in a series of transactions, if at least one transaction in the series is an avoidance transaction, the second step has been satisfied (Copthorne, at para. 64). As explained in Trustco, a series of transactions involves transactions that are “‘pre‑ordained in order to produce a given result’ with ‘no practical likelihood that the pre-planned events would not take place in the order ordained’” (para. 25, citing Craven v. White, [1989] A.C. 398 (H.L.), at p. 514, per Lord Oliver). A series of transactions also includes “related transactions or events . . . in contemplation of the series” (s. 248(10)), which refers to transactions or events before or after the series which were undertaken “‘in relation to’ or ‘because of’ the series” (Copthorne, at para. 46, citing Trustco, at para. 26).

(3)          Abusive Tax Avoidance

[56]                         The third step of the GAAR  analysis is frequently the most contentious. Indeed, it is the only step at issue in the present appeal. Analyzing whether the transactions are abusive involves, first, determining the object, spirit and purpose of the relevant provisions and, second, determining whether the result of the transactions frustrated that object, spirit and purpose (Trustco, at para. 44; Copthorne, at paras. 69‑71).

[57]                         The object, spirit and purpose reflects the rationale of the provision. The provision’s text, context and purpose help to shed light on this rationale. Once the object, spirit and purpose has been ascertained, the abuse analysis focuses on whether the result of the transactions frustrates the provision’s rationale. I provide guidance on these aspects below.

(a)           The Object, Spirit and Purpose Reflects the Rationale of the Provision

[58]                         To determine whether a transaction is abusive, courts must identify the object, spirit and purpose of the provisions alleged to have been abused, with reference to the provisions themselves, the scheme of the Act and permissible extrinsic aids (Trustco, at para. 55). The object, spirit and purpose of the provisions has been referred to as the “legislative rationale that underlies specific or interrelated provisions of the Act ” (Copthorne, at para. 69, citing V. Krishna, The Fundamentals of Income Tax Law (2009), at p. 818).

[59]                         At this juncture, it is critical to distinguish the rationale behind a provision from the means chosen to give that rationale effect within the provision. The drafting process reflects the task of translating government aims into legislative form in order to create intelligible, legally effective rules (see, e.g., Canada, Privy Council Office, Guide to Making Federal Acts and Regulations (2nd ed. 2001), at pp. 122-29). The means selected by drafters and adopted by Parliament are relevant indicia within the broader text, context and purpose analysis, since they may shed light on the rationale underlying the provision. However, the means do not necessarily provide a full answer as to why the provision was adopted (Canada v. Oxford Properties Group Inc., 2018 FCA 30, [2018] 4 F.C.R. 3, at para. 101). This is not to imply that Parliament cannot translate its aims into effective legislation — quite the opposite: when drafting legal tests, Parliament is seeking to establish a general standard that is most faithful to its objectives from the options which are available and practicable. But even the most carefully drafted provision can be abused, which is why the GAAR  exists to protect the provision’s underlying rationale.

[60]                         The object, spirit and purpose of a provision must be worded as a description of its rationale (Copthorne, at para. 69). When articulating the object, spirit and purpose of a provision, a court is not repeating the test for the provision, nor is it crafting a new, secondary test that will apply to avoidance transactions. Discerning the object, spirit and purpose does not rewrite the provision; rather, the court merely takes a step back to formulate a concise description of the rationale underlying the provision, against which a textually compliant transaction must be scrutinized (Trustco, at para. 57; Copthorne, at para. 69).

[61]                         For example, for a provision conferring a tax benefit, the rationale might relate to the basis for providing relief to taxpayers in such circumstances or, for targeted relief, the conduct that Parliament sought to encourage. Conversely, for a specific anti-avoidance rule, the rationale might relate to the specific result, or mischief, that Parliament sought to prevent.

(b)          The Provision’s Text, Context and Purpose Are Used to Determine Its Rationale

[62]                         Although the GAAR analysis involves a consideration of the provision’s text, context and purpose, the use of these elements differs from “traditional” statutory interpretation (Copthorne, at para. 70; Alta Energy, at para. 30, per Côté J., and at para. 116, per Rowe and Martin JJ., dissenting but not on this point; Oxford Properties Group, at paras. 40-44). It must be recalled that the GAAR  is a provision of last resort (Trustco, at para. 21; Copthorne, at para. 66). There is a distinction between the application of a provision in general and the application of the GAAR  to a transaction motivated by tax avoidance. If a court is performing a GAAR  analysis, the impugned transactions necessarily comply with the provisions of the Act , properly interpreted and applied (see Copthorne, at para. 88; 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. 391). This is self-evident: if there is a specific provision with which the taxpayer has not complied, the Minister need not resort to the GAAR .

[63]                         In traditional statutory interpretation, the court considers a provision’s text, context and purpose to determine what the words of the statute mean. In the GAAR analysis, however, “[t]he search is for the rationale that underlies the words that may not be captured by the bare meaning of the words themselves” (Copthorne, at para. 70 (emphasis added); Triad Gestco, at para. 51). The object, spirit and purpose analysis has a precise function: to discern the underlying rationale of the provisions. A consideration of the text, context and purpose gives structure to this analysis. Indeed, the object, spirit and purpose analysis should not turn into a “value judgment of what is right or wrong nor . . . what tax law ought to be or ought to do” (Copthorne, at para. 70). Nor should it become a “search for an overriding policy of the Act ” that is not founded in the text, context and purpose of the provisions (Canada Trustco, at para. 41; Alta Energy, at para. 49). Rather, a focus on the provision’s text, context and purpose ensures that the intrinsic and extrinsic evidence used to discern a provision’s rationale remains tied to the provision itself. To that end, a brief discussion on how to conduct this analysis is useful.

[64]                         The text of the provision is relevant to the analysis of a provision’s object, spirit and purpose (Alta Energy, at para. 58). Bearing in mind the search for the provision’s underlying rationale, courts may ask how the text sheds light on what the provision was designed to achieve. Put differently, what was the provision intended to do? (Copthorne, at para. 88). This includes considering what the text of the provision expressly permits or restricts. Similarly, the language and structure of the provision can sometimes be evocative of Parliament’s underlying concerns. The text can also help to identify the nature (or “type”) of provision at issue, which can be relevant to understanding the rationale underlying it.

[65]                         There may be circumstances where the provision’s underlying rationale is no broader than its text because, having regard to its context and purpose, the provision’s text “fully explains its underlying rationale” (Copthorne, at para. 110). A court, however, must not lose sight of the goal of the exercise — to discern the underlying rationale of the provision — and the reality that this rationale “may not be captured by the bare meaning of the words themselves” (Copthorne, at para. 70). As explained by Noël C.J. of the Federal Court of Appeal, “[w]hile one cannot rule out the possibility that the underlying rationale for a provision will be fully captured by the words, this must still be demonstrated by inquiring into the provision’s reason for being” (Oxford Properties Group, at para. 88, citing Copthorne, at paras. 110-11).

[66]                         Beyond the provision’s text, a court must consider the provision’s context. The contextual analysis “involves an examination of other sections of the Act , as well as permissible extrinsic aids” (Copthorne, at para. 91; Trustco, at para. 55). Of course, this does not involve considering every other section of the Act . Rather, “relevant provisions are related ‘because they are grouped together’ or because they ‘work together to give effect to a plausible and coherent plan’” (Copthorne, at para. 91, citing R. Sullivan, Sullivan on the Construction of Statutes (5th ed. 2008), at pp. 361 and 364).

[67]                         The focus is on the relationship between the provision alleged to have been abused and the particular scheme within which it operates (Triad Gestco, at paras. 26 et seq.). Although the Act is lengthy and detailed, an understanding of its structure can help to identify the function of the provision at issue. For example, a specific restriction may shed light on the rationale of a general benefit-conferring rule, and vice versa.

[68]                         Finally, understanding the provision’s purpose is central to the GAAR  analysis. A purposive analysis permits courts to consider legislative history and extrinsic evidence (see R. Sullivan, The Construction of Statutes (7th ed. 2022), at § 9.03, at paras. 7-8). These materials provide insight into the rationale for specific provisions. Of course, tax provisions can serve a variety of independent and interlocking purposes (Trustco, at para. 53). Nevertheless, such provisions are intended to promote particular aims, and courts must therefore determine what outcome Parliament sought to achieve through the specific provision or provisions (Copthorne, at para. 113).   

(c)          The Abuse Analysis Focuses on Whether the Result of the Transactions Frustrates the Provision’s Object, Spirit and Purpose

[69]                         At the abuse stage, the avoidance transactions will be abusive where the outcome or result of the avoidance transaction “(a) is an outcome that the provisions relied on seek to prevent; (b) defeats the underlying rationale of the provisions relied on; or (c) circumvents certain provisions in a manner that frustrates the object, spirit and purpose of those provisions” (Lipson, at para. 40, citing Trustco, at para. 45). These considerations are not independent of one another and frequently overlap (Copthorne, at para. 72). Ultimately, the analysis remains squarely focused on abuse. Courts must go beyond the legal form and technical compliance of the transactions; they must compare the result of the transactions to the underlying rationale of the provision and determine whether that rationale has been frustrated. In coming to such a conclusion, the abusive nature of the transaction “must be clear” (Trustco, at paras. 62 and 66; Copthorne, at para. 68; Alta Energy, at para. 33).

[70]                         This Court’s jurisprudence sheds light on the types of circumstances that rise to the level of abuse. For example, if the rationale underlying the provision is to encourage particular relationships or activities, abusive tax avoidance may be found where the relationships and transactions are “wholly dissimilar to the relationships or transactions that are contemplated by the provisions”, as was the case in Mathew (para. 57, citing Trustco, at para. 60). Similarly, where a specific anti-avoidance rule is flipped on its head to enable tax avoidance, there is likely to be a finding of abuse, as was the case in Lipson (para. 42; see also Fiducie Financière Satoma v. The Queen, 2018 FCA 74, 2018 D.T.C. 5052, at para. 52). An abuse may also be found in certain circumstances where a series of transactions “achieved a result the section was intended to prevent” while narrowly avoiding application of the provision, as in Copthorne (paras. 124-27). These examples are not exhaustive, but provide useful guidance on how the object, spirit and purpose of different types of tax provisions can be frustrated.

[71]                         Importantly, there is no bar to applying the GAAR  in situations where the Act  specifies precise conditions that must be met to achieve a particular result, as with a specific anti-avoidance rule. Thus, I do not agree with the appellant’s submission that where Parliament has legislated with precision, as here, where loss carryovers are denied in specific instances, the GAAR  is not meant to play a role. Of course, the GAAR  will not apply in all circumstances — the analysis is inherently case specific. Further, the way a provision has been drafted is important within the text, context and purpose analysis, since it may shed light on the conduct that Parliament sought to target and how it went about doing so. But the proposition that the GAAR  can have almost no role where Parliament has legislated a specific anti-avoidance rule is to read a restriction into s. 245 without a basis for doing so. It ignores the fact that the GAAR  was enacted in the first place partly because specific anti-avoidance rules were being circumvented through abusive tax planning, and that such rules are among those most commonly found to have been abused in GAAR  decisions (J. Li, “The Misuse or Abuse Exception: The Role of Economic Substance”, in Arnold, The General Anti-Avoidance Rule, 295, at pp. 299, fn. 25, and p. 316).

[72]                         The appellant’s position also runs contrary to this Court’s jurisprudence. As the majority recognized in Alta Energy, “[a]busive 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. 32 (emphasis added), citing Trustco, at para. 45). Moreover, this Court in Copthorne largely rejected the argument that where Parliament has drafted detailed provisions, then a taxpayer that has technically complied with these provisions cannot frustrate their rationale (paras. 108-11). Simply put, specific and carefully drafted provisions are not immune from abuse. As with any other provision, the GAAR  ensures that the rationale behind such provisions is not frustrated by abusive tax strategies.

(d)          Summary

[73]                         In summary, at the third stage of the GAAR  analysis:

-           The object, spirit and purpose is a description of the provision’s underlying rationale. The means (the how) do not always provide a full answer as to the rationale underlying the provision (the why).

-           The text, context and purpose of a provision provide indicia of its rationale. The text can shed light on what the provision was designed to encourage or prevent based on what it expressly permits or restricts, how it is worded and structured, and the nature of the provision. Similarly, the context can serve to identify the function of the provision within a coherent scheme. Finally, the provision’s purpose can help to discern the outcomes that Parliament sought to achieve or prevent.

-           Once the object, spirit and purpose has been ascertained, the abuse analysis goes beyond the legal form and technical compliance of the transactions to consider whether the result frustrates the provision’s rationale.

[74]                         Having summarized the proper approach to applying the GAAR  in light of this Court’s jurisprudence, I now turn to the transactions at issue.

VI.         Application

A.           Which Provisions Are at Issue?

[75]                         The parties do not dispute that the transactions complied with the text of the Act . The question is whether the GAAR  applies to deny the deductions and uphold the Minister’s reassessments. The appellant did not challenge the Tax Court’s findings that there was a series of transactions that resulted in a tax benefit and that they constituted avoidance transactions. Thus, the only question in dispute is whether they abused certain provisions of the Act  (the third stage of the GAAR  analysis). This requires, first, ascertaining the object, spirit and purpose of a provision and, second, determining whether it has been frustrated. Before proceeding, it is worth determining which provisions should be the focus of the analysis.

[76]                         The Minister reassessed the appellant’s claims relating to the deduction of non-capital losses, SR&ED expenditures and ITCs. However, since the restrictions on loss carryovers in each context operate in a similar way, a violation of the GAAR  in one case would lead to a similar determination in the others. This is consistent with how the parties argued the case and how it was analyzed by both the Tax Court and the Federal Court of Appeal (T.C.C. reasons, at para. 87; C.A. reasons, at para. 5).

[77]                         The courts below focused on three non-capital loss provisions: s. 111(1)(a), which sets out the benefit-conferring rule permitting non-capital loss carryovers; s. 111(5), which denies loss carryovers where the control of the corporation has been acquired by a person or group of persons, unless a continuity of business exists; and s. 256(8), which deems an acquisition of control to occur where certain share-related rights are obtained. These provisions are interrelated and complement each other, since they exist within a scheme regulating the carryover of non-capital losses. As set out below, an analysis of s. 111(5) is sufficient to dispose of the appeal.

B.            What Is the Object, Spirit and Purpose of Section 111(5)?

[78]                         The first stage of the abuse analysis, ascertaining the object, spirit and purpose of the provisions, is an extricable question of law, subject to a correctness standard of review (see Alta Energy, at para. 50, citing Trustco, at para. 44; Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 S.C.R. 235, at para. 8). To assess the underlying rationale of s. 111(5), it is necessary to consider the provision’s text, context and purpose (Copthorne, at para. 70). A review of the provision’s text, context and purpose reveals that Parliament intended to deny unused losses to unrelated third parties who take control of a corporation and change its business. I would formulate the object, spirit and purpose of s. 111(5) as follows: to prevent corporations from being acquired by unrelated parties in order to deduct their unused losses against income from another business for the benefit of new shareholders. In order to consider how this rationale flows from s. 111(5)’s text, context and purpose, I begin by analyzing the text.

(1)          The Text of the Provision

[79]                         Broadly stated, s. 111(5) is a restriction on a taxpayer’s ability to make use of its non-capital losses incurred in another taxation year. At the time of the transactions in issue, the relevant portion of s. 111(5) read as follows: 

(5) Where, at any time, control of a corporation has been acquired by a person or group of persons, no amount in respect of its non-capital loss or farm loss for a taxation year ending before that time is deductible by the corporation for a taxation year ending after that time and no amount in respect of its non-capital loss or farm loss for a taxation year ending after that time is deductible by the corporation for a taxation year ending before that time except that

(a) such portion of the corporation’s non-capital loss or farm loss, as the case may be, for a taxation year ending before that time as may reasonably be regarded as its loss from carrying on a business and, where a business was carried on by the corporation in that year, such portion of the non-capital loss as may reasonably be regarded as being in respect of an amount deductible under paragraph 110(1)(k) in computing its taxable income for the year is deductible by the corporation for a particular taxation year ending after that time

(i) only if that business was carried on by the corporation for profit or with a reasonable expectation of profit throughout the particular year, and

(ii) only to the extent of the total of the corporation’s income for the particular year from that business and, where properties were sold, leased, rented or developed or services rendered in the course of carrying on that business before that time, from any other business substantially all the income of which was derived from the sale, leasing, rental or development, as the case may be, of similar properties or the rendering of similar services . . . .

Three elements of the text warrant consideration: s. 111(5)’s reference to control; its focus on an acquisition by a person or group of persons; and the continuity of business exception. While the text of s. 111(5) has been amended since the transactions in question, the provision still encompasses these core elements, albeit structured differently (see s. 111(5.4)).

[80]                         First, the text of s. 111(5) references “control”. While control is not expressly defined within the provision or the Act , this Court concluded in Duha Printers that it refers to de jure control. The general test for de jure control is “whether the controlling party enjoys, by virtue of its shareholdings, the ability to elect the majority of the board of directors” (Duha Printers, at para. 36; R. Taylor and M.‑C. Marcil, “Duha Printers Revisited: Issues Regarding Corporate Control” (2022), 70 Can. Tax J. 495, at pp. 504-7; G. Lord et al., Les principes de l’imposition au Canada (13th ed. 2002), at s. 3.3; Explanatory Notes, at pp. 502-4; at common law, see Buckerfield’s Ltd. v. Minister of National Revenue, [1965] 1 Ex. C.R. 299). The power to select a majority of directors generally means the power to choose those who manage the affairs of the corporation, subject, in particular, to the constating documents or a unanimous shareholder agreement (Duha Printers, at para. 85; see also Canada Business Corporations Act , R.S.C. 1985, c. C-44  (“CBCA ”), ss. 2(3) and 102(1); Business Corporations Act, R.S.O. 1990, c. B.16, ss. 1(5) and 115(1); Business Corporations Act, CQLR, c. S-31.1, ss. 2 “to control” and 112; Civil Code of Québec, arts. 310, 335 and 336). I will return to the concept of de jure control, in contrast to de facto control, below.

[81]                         Second, s. 111(5) does not refer merely to a change in control, but to a situation where control has been “acquired by a person or group of persons”. The provision indicates Parliament’s focus on circumstances in which there is a shift in the locus of control, rather than, for example, high turnover in the shares of a publicly traded company (Silicon Graphics Ltd. v. Canada, 2002 FCA 260, [2003] 1 F.C. 447, at para. 36).

[82]                         Third, s. 111(5) creates an exception to the rule it sets out: losses remain deductible if, after an acquisition of control, the corporation engages in the same or a similar business. Thus, the connection to past losses is severed only when control has been acquired and there is a break from the corporation’s past business. The text of s. 111(5) reflects a concern with denying loss carryovers when there is a lack of continuity within the corporation, as measured by both the identity of its controlling shareholders and its business activity.

[83]                         The appellant argues that the object, spirit and purpose of the provision is captured by the de jure control test within s. 111(5). In some cases, the object, spirit and purpose may be no broader than the provision itself. However, this is only where the text fully explains the provision’s underlying rationale (Copthorne, at para. 110). To determine whether this is the case for s. 111(5), the analysis must move to the context and purpose of the provision.

(2)          The Context of the Provision

[84]                         It would be difficult to understand the role of s. 111(5) within the Act  without considering the function it plays in the larger scheme of computing taxable income. Section 111(5) operates as a step within this broader process.

(a)          Section 111(5) Should Be Considered Against the Foundational Principles of the Act

[85]                         The roots of the basic income tax system of the Act  lie in ss. 2 to 4 (Li (2021), at p. 307). Section 2(1)  of the Act  provides that income tax shall be paid “on the taxable income for each taxation year of every person resident in Canada”. The Act  treats each “person” as a tax unit; every “person”, including a corporation, is a separate taxpayer. While this section is general in nature, it reflects the foundational principle that “taxpayers are to be taxed on their own earnings, and not the earnings of someone else” (Canada v. 594710 British Columbia Ltd., 2018 FCA 166, [2019] 5 C.T.C. 1, at para. 54). Consequently, the effect of s. 111(5) becomes more clear when read in light of s. 2: s. 111(5) suggests that, at a conceptual level, when there has been an acquisition of control, and the corporation’s loss-generating business ceases to operate, the corporation can no longer be understood as the same taxpayer for the purpose of carrying over losses from the corporation’s prior operations.

(b)          Section 111(5) Delineates the Boundaries of the Benefit-Conferring Provision, Section 111(1)(a)

[86]                         Section 111(5) operates within a scheme for carrying over non-capital losses. The foundational provisions of the Act  recognize that each taxpayer is taxed based on their income and losses within a single taxation year; however, s. 111(1)(a) modifies this general rule and provides that a taxpayer can deduct non-capital losses against income in a future or prior taxation year. It served as the benefit-conferring provision which the appellant relied on to obtain its tax benefits. Without such a loss carryover rule, a taxpayer who earns high income in one year and suffers a loss in the next would face financial hardship as a result of paying tax in the first year without relief in the second (Krishna, at p. 595). The taxpayer with fluctuating income is disadvantaged compared with the taxpayer who engages in a less risky venture, paying tax over time on a more stable income.

[87]                         From this perspective, s. 111(5) does not operate independently, but rather serves to complement s. 111(1)(a). There is sound logic behind the ability to carry over one’s losses. However, the specific purpose of s. 111(1)(a) has always been to more accurately measure income and provide relief to the individual taxpayer, accounting for fluctuating income over a period of multiple years. Amendments to this provision have broadened the ability to deduct non-capital losses (see, e.g., S.C. 1958, c. 32, s. 12(1), amending s. 27(1)(e) of the Income Tax Act, R.S.C. 1952, c. 148), but never so as to benefit a new taxpayer who did not themselves generate the losses. Only the taxpayer who suffered the loss is entitled to deduct the loss.

[88]                         Section 111(5) ensures that this principle is given effect for corporations. While the corporation is still the same legal person after an acquisition of control, the identity of those behind the corporation has changed. Section 111(5) functions so that the tax benefits associated with those losses will not benefit a new shareholder base carrying on a new business.

[89]                         This restriction on s. 111(1)(a) promotes consistency with other provisions within the scheme for computing income tax. In particular, s. 249(4) specifies that when an acquisition of control occurs, the corporation’s taxation year is deemed to have ended and a new taxation year commences. Thus, an acquisition of control severs the link between the corporation’s prior and post-acquisition operations. It would be incongruous with several other provisions if the Act  nevertheless allowed this corporation to carry forward past losses as if the link had not been severed.

[90]                         Thus, s. 111(5) serves to delineate the boundaries of s. 111(1)(a) and to promote consistency with other provisions which treat the corporation as, effectively, a new taxpayer following an acquisition of control (Li (2021), at p. 314; Strain, Dodge and Peters, at pp. 4:52-54).

(c)          Parliament’s Selection of Control Tests Differs Across the Act

[91]                         The appellant places significant emphasis on s. 256(5.1)  of the Act , which defines the de facto control test. This test was adopted after the introduction of s. 111(5) and while numerous provisions were amended to either adopt or reject the de facto control test, the de jure control test in s. 111(5) was left unchanged. The appellant argues that the use of the de facto control test in other provisions indicates that “Parliament turned its mind to the question of control throughout the Act  . . . Parliament understood the difference between de facto and de jure control and intended that difference” (A.F., at para. 64 (emphasis in original)). For the appellant, this is effectively determinative of the dispute, and reduces the object, spirit and purpose of s. 111(5) to the de jure control test.

[92]                         In my view, the context of the Act  reveals that, when faced with a choice between de jure and de facto control as the general test, Parliament opted for de jure control. There are various reasons why Parliament would have chosen the de jure control test as the standard to be used on an application of s. 111(5), something this Court addressed in Duha Printers, at para. 58:

In my view, the de jure standard was chosen because in some respects it is a relevant and relatively certain and predictable concept to employ in determining control. In general terms, de jure refers to those legal sources that determine control: namely, the corporation’s governing statute and its constitutional documents, including the articles of incorporation and by-laws. The de facto concept was rejected because it involves ascertaining control in fact, which can lead to a myriad of indicators which may exist apart from these sources. [Emphasis in original.]

[93]                         There is no doubt that de jure control is a clearer benchmark than de facto control, meaning greater certainty for the majority of transactions, which are not tax-motivated. De facto control, by contrast, is a broad test that can be met by mere indirect influence, and therefore does not require the existence of legally enforceable rights (s. 256(5.1); Krishna, at p. 695). Such a test would have captured a variety of situations far beyond Parliament’s concern for ensuring that one taxpayer would not benefit from another’s losses. For example, there may exist significant continuity in the corporation’s directors and shareholders, but an onerous loan designed to help a distressed corporation reorient its business could create a relationship of dependence sufficient to trigger the de facto control test, even if the creditor’s influence was never exercised. Thus, when deciding between two imperfect tests, it is logical that Parliament would opt for de jure control.

[94]                         However, it does not follow that the provision’s rationale is fully captured by the de jure control test. Rather, de jure control was the marker that offered a roughly appropriate proxy for most circumstances with which Parliament was concerned — particularly given that the GAAR  exists as a last resort. To ascertain the rationale underlying s. 111(5), more is needed than the simple fact that Parliament settled on this test to operationalize its intent.

[95]                         Indeed, the rationale of s. 111(5) is illuminated by related provisions which both extend and restrict the circumstances in which an acquisition of control has occurred. These provisions suggest that de jure control is not a perfect reflection or complete explanation of the mischief that Parliament sought to address.

(d)          The Control Test in Section 111(5) Is Expanded and Restricted by Other “Deeming” Provisions

[96]                         Section 256(8) deems control to have been acquired when, for the purpose of avoiding certain provisions, a taxpayer obtains a right to, inter alia: acquire shares; cause a corporation to redeem, acquire or cancel shares; acquire or control voting rights; or cause the reduction of other shareholders’ voting rights (see s. 251(5)(b)). If the listed right, when exercised, would provide the taxpayer with de jure control, s. 256(8) deems control to have been acquired when the right was acquired. This subsection is notable because, although it remains focused on share-related rights, it looks beyond the standard documentation considered under the de jure control test. Instead, it assesses legal rights, including contractual agreements. Section 256(8) closes a loophole where corporations could have acquired a contractual right that, if exercised, would have enabled them to exert control over the corporation (Fiscal Statement (1980), at p. 40; G. W. Flynn, “Tax Planning for Corporations with Net Capital and Noncapital Losses”, in Corporate Management Tax Conference, Current Developments in Measuring Business Income for Tax Purposes (1982), 208, at p. 209). More broadly, it demonstrates that the de jure control test in s. 111(5) does not, on its own, capture the full range of situations that Parliament sought to target; rather, the test in s. 111(5) is better understood as a rough proxy that seeks to give effect to Parliament’s broader aims, while offering a degree of clarity and stability in most cases.

[97]                         By contrast, s. 256(7) deems control not to have been acquired in certain circumstances in order to give proper effect to provisions similar to s. 111(5). For example, it ensures that transfers can occur between related parties without the unused losses being denied due to a “technical change in control of the corporation” (A. R. A. Scace and D. S. Ewens, The Income Tax Law of Canada (5th ed. 1983), at p. 106; s. 256(7)(a) and (b)). This suggests that the identity of those who have acquired control matters when determining whether the corporation should be treated as a new taxpayer. In this way, s. 111(5) and related provisions are not merely concerned with a formal change of control: if the corporation has not changed hands, the Act  recognizes the continuity between its past losses and future business endeavours. It is only when an unrelated party takes the reins that permitting loss carryovers would undermine the fundamental principles underpinning the Act and the integrity of the loss carryover rule in s. 111(1)(a).

[98]                         I note that s. 256.1, enacted after the transactions at issue, includes a deeming rule for substantial equity acquisitions. However, consideration of a provision enacted subsequent to the transactions at issue is neither necessary nor warranted in this case (see also Oxford Properties Group, at paras. 46 and 86). 

[99]                         The foregoing contextual analysis sheds significant light on the rationale behind s. 111(5). However, to more accurately ascertain its object, spirit and purpose, it is necessary to consider extrinsic evidence of Parliament’s purpose throughout the legislative history of the provision.

(3)          The Purpose of the Provision

[100]                     The legislative history behind s. 111(5) illustrates that Parliament was concerned with addressing the trading of loss corporations, which was undermining the tax base and creating inequity among taxpayers. This reflects a specific concern that fits within the Act ’s more general policy of prohibiting the transfer of losses between taxpayers, subject to specific exceptions (Mathew, at para. 53). 

[101]                     Importantly, a purposive analysis of s. 111(5) also demonstrates that, while the means Parliament has chosen to address these concerns has evolved over time — finally settling on de jure control — its rationale for including the non-capital loss carryover restriction in the Act  has been consistent.

[102]                     The legislative history of s. 111(5) begins in 1958. In 1958, s. 27(1)(e) of the Income Tax Act, R.S.C. 1952, c. 148 — the precursor to s. 111(1)(a) — was amended “to permit business losses to be applied to reduce income not only from the same business, but also from any other business carried on by the taxpayer” (Strain, Dodge and Peters, at p. 4:42). This was a significant liberalization of the loss carryover rules; previously, business losses were deductible only against income generated in other years from the same business.

[103]                     In the same amendment, s. 27(5) — the precursor to s. 111(5) — was added to the Act . This subsection read as follows:

(5) Paragraph (e) of subsection (1) does not apply to permit a corporation to deduct, for the purpose of computing its taxable income for a taxation year, a business loss sustained by it in a preceding taxation year, in any case where

(a) more than 50% of the shares in the capital stock of the corporation have, between the end of that preceding year and the end of the taxation year, been acquired by a person or persons who did not, at the end of that preceding year, own any of the shares in the capital stock of the corporation; and

(b) the corporation was not, during the taxation year, carrying on the business in which the loss was sustained.

 

(S.C. 1958, c. 32, s. 12(2))

[104]                     Commentary in the Senate explained that without such a restriction, “it would be quite easy to purchase companies that have incurred losses and then get the benefit” of the carryforward (Debates of the Senate, 1st Sess., 24th Parl., August 21, 1958, at pp. 650-51 (Hon. Gunnar S. Thorvaldson)). Put another way, the restriction was enacted “for the purpose of checking the practice of trading in tax-loss corporations” (H. D. McGurran, “Principles of Income Tax: 2. Taxable Income from a Business” (1958), 6 Can. Tax J. 455, at p. 465). At the same time, the restriction on business loss carryovers introduced by s. 27(5) could be avoided if the business in which the loss was sustained continued to be carried on (i.e., a continuity of business exception). Notably, the test Parliament adopted at the time (more than 50 percent of the shares in the corporation’s capital stock, acquired by one or more new shareholders in a single year) was distinct from the current marker (de jure control).

[105]                     After s. 27(5) was enacted, commentators were quick to identify problems and loopholes. One author, for example, described how the new carryover restriction could be circumvented with only a bit of foresight:

If a prospective buyer had the foresight to acquire a nominal number of shares, or for that matter, one share, before the turn of the fiscal year, then the full amount of the losses would be preserved regardless of what type of business was in future carried on. Perhaps negotiations for the purchase of loss companies are best conducted in the final weeks of their fiscal year.

 

(C. W. Leach, “Making the Most of Your Losses”, in Canadian Tax Papers No. 19, Corporate Management Conference (1960), 27, at p. 29)

[106]                     In 1963, s. 27(5)(a) was amended to refer to an acquisition of control of a corporation by a person or persons who did not control the corporation at the end of the preceding year, while still subjecting transactions that predated the amendment to the former test relating to an acquisition of 50 percent of its shares (S.C. 1963, c. 21, ss. 6(1) and (2), substituting s. 27(5)(a) and adding s. 27(5a)). At the Bill’s Second Reading in the House of Commons, then Minister of Finance Walter L. Gordon raised concerns about the trafficking in shares of loss companies:

The public, including the business public, tends to lose respect for any tax administration when advantage can be taken freely of obvious loopholes in the law . . . it seems desirable to stop some clear opportunities for tax avoidance, and I should like to mention four examples of what it is intended to do about them.

 

. . .

 

The second example that I would like to mention is the one that at present provides an opportunity, under the act, for companies which incur losses in one year to carry them forward and apply them against profits earned during subsequent years.

 

As a result of this, a practice has developed of trafficking in the shares of companies whose businesses have been discontinued, but which technically have certain tax loss carry forward entitlements. For example, a man with a profitable business may purchase the shares of one of these companies and transfer his prosperous business to it, and having done so he can apply the losses incurred previously, and often in a quite different class of business, against the taxable income of his own business. I suggest that surely this was never the intention of parliament when the original proposal was incorporated in the act, and that this practice should be stopped. [Emphasis added.]

 

(House of Commons Debates, vol. IV, 1st Sess., 26th Parl., October 16, 1963, at pp. 3636-37 (Hon. Walter L. Gordon))

[107]                     The Minister of Finance later made the following statement about the provision’s purpose:

The purpose of this clause is to prevent a company with a loss carry forward entitlement from changing hands, thus permitting a new owner to introduce into that business an entirely different business after the date of the budget, and having this entitlement applied against the profits of the new business enterprise. . . .

 

. . .

 

. . . This will wipe out the loss credit position if there has been a change in ownership. The reason, of course, is obvious. People with profitable businesses who were paying taxes on their profits were making a practice of acquiring companies, often shells which had nothing in them but which were in possession of a loss-carry-forward entitlement, for the purpose of applying this entitlement against their profits and thus avoiding the payment of tax. That is what this clause is intended to prevent. I am prepared to admit that in theory, certainly, if you look upon the corporation which has the entitlement as being a separate entity, it does lose its rights, where its ownership has been transferred to somebody else. However, that is the very thing we are trying to stop. [Emphasis added.]

 

(House of Commons Debates, vol. V, 1st Sess., 26th Parl., November 1, 1963, at p. 4287 (Hon. Walter L. Gordon))

[108]                     In 1966, the Carter Commission examined the Income Tax Act  and made a variety of recommendations that were then considered by Parliament in subsequent years. It sought to explain the purpose for the restriction on loss transfers at a conceptual level:

Under the present tax system, the new owner of an unincorporated business does not obtain any deduction in respect of unabsorbed losses of the previous owner. The same position arises where assets of an incorporated company with unabsorbed losses are purchased. In each of these cases the purchaser is a different taxpayer from the vendor and is not permitted to utilize the losses of the vendor for tax purposes. 

 

. . .

 

We have stated our belief that a corporation should be regarded as an intermediary for the shareholders. The proposed liberal allowance of losses is not intended to be used in such a way that one taxpayer can deduct losses sustained by another and thereby defer or avoid tax liability. Accordingly, we recommend that losses should not be transferable from one taxpayer to another, and that the right to carry losses forward should be denied to a corporation where there is a change in control, either through a sale of its shares, through granting a right to acquire a controlling interest (unless the right is exercisable only on death or default of an obligation or under a first refusal arrangement) or through a statutory amalgamation. [Emphasis added.]

 

(Canada, Report of the Royal Commission on Taxation (1966), vol. 4, at pp. 261-62)

[109]                     In 1972, s. 27(5) was renumbered as s. 111(5). It continued to limit carryovers in the case of an acquisition of control and continued to provide a continuity of business exception. The “50% of the shares” test was dropped, and s. 111(5) now referred only to an acquisition of control (S.C. 1970-72, c. 63, s. 1). The carryover restrictions were tightened through subsequent amendments (Strain, Dodge and Peters, at p. 4:42); however, this has remained the test since.

[110]                     The foregoing legislative record demonstrates that, while Parliament has had to “redefine terms in order to enforce its intention” (G. W. Riehl, Incorporation and Income Tax in Canada (4th ed., 1965 at p. 69), the provision has always restricted non-capital loss carryovers upon a change in the locus of control of a corporation from one person or group to another, subject to a continuity of business exception. When a corporation changes hands, and the loss business ceases to operate, the corporation is effectively a new taxpayer who cannot avail itself of non-capital losses accumulated by the old taxpayer. The following provides a useful explanation:

First and foremost, the carryover of losses following a change of control is not generally supported in tax policy terms. Normally, one taxpayer cannot avail himself of another taxpayer’s losses. In the case of an artificial entity such as a corporation, when its control changes it is essentially regarded as a new taxpayer, because different shareholders then become entitled indirectly to enjoy the benefits of its financial success. [Emphasis added.]

 

(Strain, Dodge and Peters, at p. 4:52)

[111]                     The business continuity exception does not undermine this conclusion: while this rule may be viewed as arguably benefitting new shareholders, there is a reason for the exception. It encourages “the recovery of unprofitable enterprises” (Strain, Dodge and Peters, at p. 4:53), which may require a new investment by new owners. Where this occurs, “the takeover has accomplished the sound commercial objective of making an unprofitable business profitable, and there is no reason why the unused pre-takeover losses should not continue to be available” (P. W. Hogg, J. E. Magee and T. Cook, Principles of Canadian Income Tax Law (3rd ed. 1999), at p. 408, cited with approval in OSFC Holdings Ltd. v. Canada, 2001 FCA 260, [2002] 2 F.C. 288, at para. 91). Indeed, it suggests that those who have taken the reins were interested in strengthening the corporation’s business, rather than using the corporation as a vessel for unrelated activities that would have distorted its identity. Although the corporation may have changed hands, the link in continuity is preserved through a different marker: the corporation’s business remains consistent. Moreover, the justification for s. 111(1)(a) — to more accurately measure income and provide relief to an individual taxpayer with fluctuating income over time — remains applicable, given this continuity. This reinforces that, at its core, s. 111(5) serves to delineate the circumstances in which the basis for the loss carryover rule in s. 111(1)(a) is non-existent.

[112]                     In sum, the legislative history behind s. 111(5) illustrates that Parliament’s intended purpose was to address a specific mischief. While the means Parliament has chosen to address these concerns have evolved over time, the rationale for including the loss carryover restriction in the Act  has remained consistent.

(4)          Conclusion on Object, Spirit and Purpose

[113]                     In light of the foregoing, the object, spirit and purpose of s. 111(5) is to prevent corporations from being acquired by unrelated parties in order to deduct their unused losses against income from another business for the benefit of new shareholders. Parliament sought to ensure that a lack of continuity in a corporation’s identity was accompanied by a corresponding break in its ability to carry over non-capital losses. This is the rationale underlying the provision and properly explains why Parliament enacted s. 111(5).

[114]                     In articulating the object, spirit and purpose of s. 111(5), the Tax Court focused on the concept of “effective control” (para. 134), a term briefly employed by Iacobucci J. in Duha Printers (para. 36). The Federal Court of Appeal acknowledged that this wording was unclear, and replaced it with the term “actual control” (paras. 72-73). This arguably exacerbated the problem, and led the appellant to argue that the Federal Court of Appeal had substituted Parliament’s de jure control test with an ambiguous de facto control test. The appellant proposed a formulation of the object, spirit and purpose centred on de jure control.

[115]                     Respectfully, both the lower courts and the appellant formulated the object, spirit and purpose as a legal test, rather than summarizing the rationale of the provision. This ultimately distorted their GAAR  analysis. De jure control, “effective” control and “actual” control do not indicate why Parliament was concerned about an acquisition of control and the mischief it sought to address (Oxford Properties Group, at para. 101). To define the object, spirit and purpose of s. 111(5) based on Parliament’s choice of test or substitute it for another test would, in this case, result in prioritizing the means (the how) over the rationale (the why).

[116]                     In s. 111(5), Parliament has clearly chosen a test for control: de jure control. De jure control was a reasonable marker for the situations in which a corporation’s identity has changed. This being so, it is primarily a means of giving effect to Parliament’s aim, rather than a complete encapsulation of the aim itself. As with any provision, Parliament had to select a general test for s. 111(5) among the options available: it had good reason to select the de jure control test over a broad de facto control test, which would have captured a variety of conduct unrelated to Parliament’s aims. Again, the test to apply on the straightforward application of s. 111(5) is not in dispute. However, de jure control does not, in itself, explain what was concerning to Parliament, as is amply demonstrated by a careful analysis of the intrinsic and extrinsic evidence. Respectfully, my colleague’s reasons conflate the means found within a provision’s text (in this case, de jure control) with the provision’s underlying rationale; this approach would have implications for a variety of provisions involving a control test, such that the GAAR  would effectively not apply.

[117]                     Before proceeding, I make two further comments. First, accurately setting out the object, spirit and purpose that underlies the provision does not change the applicable test within the provision. Until Parliament legislates otherwise, de jure control remains the standard for the application of s. 111(5). As I explained, within the GAAR  analysis, courts are not formulating a new legal test when they determine a provision’s object, spirit and purpose; rather, they seek to capture a provision’s rationale, thereby providing a synthesis of what the provision was designed to achieve or prevent. Nor are courts “applying” the object, spirit and purpose as if it was a new bright-line test in the abuse analysis. The analysis is comparative: for a provision to be abused under a GAAR  analysis, the result that the transactions achieved — transactions which have already been shown to have the primary purpose of avoiding taxes — is assessed against the provision’s rationale to determine whether this rationale is frustrated (Trustco, at para. 57; Copthorne, at para. 69).

[118]                     In this case, s. 111(5) demonstrates Parliament intended to deny unused losses to unrelated third parties who take the reins of a corporation and change its business. This explains what Parliament sought to prevent through the provision, understood in light of its text, context and purpose.

[119]                     Second, I note that contrary to my colleague’s suggestion, this case does not resemble the dispute in Alta Energy, which concerned a carve-out provision in an international treaty that allocated to a person’s residence state the right to tax certain capital gains for immovable property. The majority determined that through the carve-out, Canada was deliberately seeking to “forego its right to tax” certain capital gains because it wanted to “encourage investments by Luxembourg residents in business assets embodied in immovable property located in Canada” (paras. 6 and 89). The rationale for this incentive applied whether or not Luxembourg residents had sufficient economic connections or were merely residents based on formal criteria: Luxembourg was well-known for its international tax haven regime, and evidence demonstrated that Canada’s “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” (para. 87). By contrast, in this case, Parliament did not make a choice to permit loss-trading strategies, nor viewed this as an acceptable tradeoff. The existence of s. 111(5), along with Parliament’s continued amendment of related provisions to extend its reach, reflects that Parliament did not intend for such a result and sought to address conduct that bore no relation to aim of the non-capital loss carryover, as is shown by an analysis of s. 111(5)’s text, context and purpose.

[120]                     The sole question in this case is whether the result of the particular series of transactions at issue is inconsistent with the rationale underlying s. 111(5). It is to this question that I now turn.

C.            Was There an Abuse of Section 111(5)?

[121]                     The second stage of the abuse analysis, determining whether the transactions are abusive, is “necessarily fact-intensive” (Trustco, at para. 44; Oxford Properties Group, at para. 39). Thus, deference is owed to the Tax Court’s factual findings (Housen, at para. 10). However, no deference is owed to the overall conclusion that there was no abuse in this case, as that conclusion was premised on an incorrect construction of s. 111(5)’s object, spirit and purpose, and therefore, was based on an extricable error of law (Trustco, at para. 46). The trial judge’s factual findings regarding the transactions themselves continue to merit deference, absent palpable and overriding error; that being said, facts which were decisive when answering the wrong legal question are not necessarily as salient when answering the right one. Thus, these findings may be more or less relevant when the transactions are considered in light of the correct object, spirit and purpose.

[122]                     An analysis of the transactions at issue demonstrates that their result served to frustrate the object, spirit and purpose of s. 111(5): considering the circumstances as a whole, they achieved the outcome that Parliament sought to prevent and provided Matco with the benefits of an acquisition of control, all while narrowly circumventing the application of s. 111(5). I note that finding that a transaction falling short of de jure control has abused s. 111(5) in this case does not mean that every transaction falling short of de jure control will be found abusive of this provision. Indeed, the question is not whether Matco holds de jure control or satisfies some other test such as de facto control. As I explained, the abuse analysis is comparative: it asks courts to assess the transactions at issue in light of the provision’s rationale to determine whether the result achieved by the transactions frustrates this rationale. It clearly did so here.

[123]                     The conclusion that the series of transactions involving the appellant were abusive flows directly from a proper consideration of what s. 111(5) was designed to achieve and how the appellant circumvented that outcome. In the analysis that follows, I consider the result obtained by the transactions, with particular emphasis on (1) the fundamental transformation of the appellant’s identity and (2) the rights and benefits obtained by Matco.

[124]                     Section 111(5)’s rationale is to prevent corporations from being acquired by unrelated parties in order to deduct their unused losses against income from another business for the benefit of new shareholders. As previously explained, s. 111(5) reflects the proposition that when the identity of the taxpayer has effectively changed, the continuity at the heart of the loss carryover rule in s. 111(1)(a) no longer exists. From this perspective, the same result was achieved through the impugned transactions. Indeed, the reorganization transactions resulted in the appellant’s near-total transformation: its assets and liabilities were shifted to Newco, such that all that remained were its Tax Attributes. Put differently, the appellant was gutted of any vestiges from its prior corporate “life” and became an empty vessel with Tax Attributes.

[125]                     Importantly, the tax attributes of a loss corporation were preserved to benefit another party. Indeed, the payments agreed upon by Matco under the Investment Agreement were in contemplation of being able to financially benefit from the Tax Attributes by filling the vessel with its chosen corporate opportunity and remaining a significant equity shareholder. Such a conclusion is reinforced by the adjustment clause in the Investment Agreement. The amounts owed by Matco under the Investment Agreement were to be adjusted based on the actual value of the Tax Attributes; in this way, the consideration which Matco was to provide was tied to its ability to benefit from monetizing the Tax Attributes (arts. 2.3, 5.5(a) and 5.5(b)).

[126]                     Moreover, the shareholder base of the taxpayer underwent a fundamental shift throughout the transactions: the appellant began as a wholly owned subsidiary of Newco and ultimately became a publicly traded company advertised for a new business, with Matco as part of the new shareholder base.

[127]                     As for its business activity, the appellant was used as the vessel for an unrelated venture planned by DKCM and selected by Matco. Thus, the only link between the appellant after the transactions and its prior corporate “life” was the Tax Attributes; in other respects, it was, in practice, a company with new assets and liabilities, new shareholders and a new business. Accordingly, the transactions resulted in a fundamental change in the identity of the taxpayer. The appellant’s continued ability to benefit from the loss carryover deductions frustrates the rationale behind Parliament’s decision to sever the continuity of tax treatment in s. 111(5), particularly considering the rights and benefits obtained by Matco.

[128]                     As I explained, the de jure control test was used as a means to implement Parliament’s aims in s. 111(5) because it appropriately recognizes that obtaining majority voting shares carries with it the ability to elect the board of directors and therefore to control the management of the affairs of the corporation (Buckerfield’s, at pp. 302-3; Duha Printers, at paras. 35-36). Matco achieved the functional equivalent of such an acquisition of control through the Investment Agreement, while circumventing s. 111(5), because it used separate transactions to dismember the rights and benefits that would normally flow from being a controlling shareholder. Several aspects of the transactions at issue demonstrate this functional equivalence, by which I mean that Matco achieved an outcome that Parliament sought to prevent without directly acquiring the rights that would have triggered s. 111(5) (Trustco, at para. 57; Copthorne, at para. 69).

[129]                     First, Matco contracted for the ability to select the corporation’s directors. Section 3.4 of the Investment Agreement reads as follows:

3.4 Resignation of Officers and Directors

 

The Parent shall exercise its best efforts to ensure that a nominee of Matco is added to the board of directors of the Subsidiary and that such nominee, as well as Charles Butt and David Goold, shall be the sole officers and directors of the Subsidiary, effective as of the Effective Time. The Parent shall exercise its best efforts to cause such directors to continue as directors of the Subsidiary until the closing of a Corporate Opportunity, at which time it shall exercise its best efforts to cause each of Charles Butt and David Goold to resign as an officer and director and provide an unconditional mutual release of all claims he or she may have against the Subsidiary and vice versa, other than claims in respect of indemnification.

 

(A.R., vol. II, at p. 87)

[130]                     Accordingly, even without holding majority voting shares, Matco contractually required Newco, who would have control over most remaining shares at the relevant time, to ensure that the three directors of the appellant would be Butt, Goold and a director selected by Matco. This is what occurred during the election following the execution of the Investment Agreement. It thereby contractually oversaw the makeup of the board of directors. As contemplated by s. 3.4, once the corporate opportunity was found and accepted, Butt and Goold were to resign and were to be replaced with directors selected by DKCM to implement its own business activity while monetizing the appellant’s Tax Attributes. Again, this is precisely what occurred.

[131]                     Second, the Investment Agreement in effect placed severe restrictions on the powers of the board of directors. The appellant was prohibited from taking a variety of actions including, but not limited to: issuing any shares, options, warrants, calls, conversion privileges or rights to acquire any shares; changing or amending its constating documents or by-laws; reorganizing, amalgamating, merging or continuing the appellant with any other legal person; taking any action relating to the liquidation, dissolution or winding-up of the appellant; declaring any dividends; entering into any contract in respect of the appellant; and engaging in any activity other than examining and pursuing a corporate opportunity (s. 6.1). Several of these decisions are so central to directors’ duties that under its governing corporate law statute, directors are not ordinarily permitted to delegate such authority to a managing director or a committee (s. 115(3)  of the CBCA ). The aforementioned corporate activities could only occur with the “prior written consent of Matco” (s. 6.1 of the Investment Agreement).

[132]                     The restrictions in favour of Matco resemble the fettering of discretion that would normally occur through a unanimous shareholder agreement and which would lead to an acquisition of de jure control (Duha Printers, at para. 71). The only reason that s. 111(5) was not violated was due to a circuit-breaker transaction: the managing director of Matco purchased 100 shares through a holding company, 1250280 Alberta Ltd., which was not a party to the agreement in order to prevent the Investment Agreement from being signed by all shareholders. Despite this manoeuvre, the result was similar: the powers of the directors were effectively neutralized for the duration of the agreement.

[133]                     Third, the transactions allowed Matco to reap significant financial benefits. Through the transactions at issue, Matco became a significant equity owner and maintained a stake in the corporation worth $4.5 million following the IPO. However, rather than merely purchasing a majority of the voting shares, it used a separate contractual agreement to gain the functional equivalent of majority voting power prior to the fulfilment of the corporate opportunity. During this intervening period, Matco deprived Newco, the majority voting shareholder on paper, of each of the core rights that it could ordinarily have exercised. One need only examine the three traditional shareholder rights outlined in s. 24(3)  of the CBCA : the right to vote at any meeting of shareholders, the right to receive any dividend declared by the corporation, and the right to receive a portion of the remaining property of the corporation on dissolution. While Newco maintained its voting rights in theory, most decisions that would normally be subject to a shareholders vote — such as a change in the corporation’s by-laws — could only occur with the consent of Matco (s. 6.1(b)(iii) of the Investment Agreement). Similarly, it may have had a right to dividends on paper, but any such dividends could only be declared with the consent of Matco (s. 6.1(c)(iv)). Finally, Newco may have been entitled to the appellant’s remaining property upon dissolution, but again, the directors were prohibited from taking such a step without Matco’s approval (s. 6.1(c)(iii)). In any case, all of the appellant’s assets had been removed through the reorganization transactions. In this way, the appellant corporation had changed hands, but Matco obtained such a result through a series of transactions rather than through the acquisition of a majority of the voting shares in the appellant.

[134]                     In response, the appellant argues that it remained at all times a free actor. In my view, any residual freedom under the Investment Agreement is illusory and merely reinforces how the transactions frustrated the rationale of s. 111(5).

[135]                     While the appellant could, in theory, refuse the corporate opportunity presented by Matco (s. 4.1 of the Investment Agreement), a careful reading of the Investment Agreement indicates that its acceptance of such an opportunity was a fait accompli. 

[136]                     First, the appellant was prohibited from engaging in any other activity other than studying and accepting the corporate opportunity (s. 6.1(d)). Put simply, the sole purpose for the appellant’s continued existence was to serve as a vessel for the corporate opportunity selected by Matco. It was therefore not in doubt that such an opportunity would be accepted.

[137]                     Second, the consequences of refusing the corporate opportunity were severe. In particular, Matco would no longer be obligated to provide the Guaranteed Amount (s. 5.5(d)). It is worth placing these consequences in context: when the appellant sought an external opportunity, it was in dire financial stress. The money owed by Matco under the Investment Agreement, in consideration for permitting it to indirectly monetize the Tax Attributes, provided the appellant with its only lifeline. The discretion provided in s. 4.1 must be understood in light of these circumstances, which were known to the parties at the time of contracting. Indeed, they are the reason that the negotiations occurred in the first place, because the appellant was unable to benefit from the Tax Attributes on its own due to its financial situation.

[138]                     As for the fact that Newco could sell its shares to another party if it received a higher price than Matco (ss. 5.1 and 5.2), the prospect of finding another buyer was illusory. Given that the appellant was unable to carry out any activities without Matco’s consent due to the Investment Agreement, it could not reasonably be expected that a third party would value the shares higher than the price stipulated within the agreement until Matco had first found a corporate opportunity.

[139]                     Of course, Newco had the right not to sell its remaining shares at all (ss. 5.3 and 5.8), as the Tax Court recognized (paras. 159 and 164). But understood in light of the correct object, spirit and purpose, the fact that Newco could choose to keep its voting shares is consistent with the circuitous path by which the transactions circumvented s. 111(5). Regardless of whether Newco sold its shares to Matco, the actions of the appellant were already locked down by the Investment Agreement, and the core benefits of share ownership (such as the right to dividends) were negated by being subjected to Matco’s approval. Furthermore, the ability to receive the Guaranteed Amount directly rather than by selling the shares to Matco (ss. 5.3 and 5.8) merely reinforces that the transactions were designed to circumvent s. 111(5). This option was important because in certain circumstances, Matco’s purchase of the shares might lead to an acquisition of de jure control. The complex series of transactions and the flexibility built into the Investment Agreement were necessary only because the contracting parties sought to achieve the very mischief that s. 111(5) was intended to prevent.

[140]                     Considering the foregoing circumstances as a whole, the result obtained by the transactions clearly frustrated the rationale of s. 111(5) and therefore constituted abuse. The object, spirit and purpose of s. 111(5) is to prevent corporations from being acquired by unrelated parties in order to deduct their unused losses against income from another business for the benefit of new shareholders. The transactions achieved the very result s. 111(5) seeks to prevent. Without triggering an “acquisition of control”, Matco gained the power of a majority voting shareholder and fundamentally changed the appellant’s assets, liabilities, shareholders and business. This severed the continuity that is at the heart of the object, spirit and purpose of s. 111(5).

VII.      Conclusion

[141]                     For the foregoing reasons, I would dismiss the appeal with costs.

                   The following are the reasons delivered by

 

                   Côté J. —

I.               Overview

[142]                     This case is of profound concern to Canadian taxpayers. The legal issue before this Court is whether the investment agreement between Matco Capital Ltd., 0813361 B.C. Ltd. (“Newco”) and the appellant, Deans Knight Income Corporation (“Investment Agreement”), amounts to an abuse of s. 111(5)  of the federal  Income Tax Act , R.S.C. 1985, c. 1 (5th Supp .) (“Act ”), for the purposes of the general anti‑avoidance rule (“GAAR ”) in s. 245  of the Act .

[143]                      It is essential to keep in mind that the GAAR  requires a careful balance “between two competing interests: the interest of the taxpayer in minimizing his or her taxes through technically legitimate means and the legislative interest in ensuring the integrity of the income tax system” (V. Krishna, Fundamentals of Canadian Income Tax (2nd ed. 2019), vol. 1, at p. 96). As Binnie J., dissenting, aptly warned in Lipson v. Canada, 2009 SCC 1, [2009] 1 S.C.R. 3, at para. 55, “[t]he GAAR  is a weapon that, unless contained by the jurisprudence, could have a widespread, serious and unpredictable effect on legitimate tax planning”. With respect, I am of the view that both my colleague Rowe J. and the Federal Court of Appeal have failed to apply the GAAR  with due restraint. Accordingly, I would allow the appeal.

[144]                     I disagree with several aspects of my colleague Rowe J.’s analysis. Despite Parliament’s unambiguous adoption of the de jure control test in s. 111(5)  of the Act , my colleague has opted for an ad hoc approach that expands the concept of control based on a wide array of operational factors. In my view, this approach invites the exercise of unbounded judicial discretion and will result in the loss‑trading restrictions in s. 111(5) being applied to transactions on a circumstantial basis. Moreover, unlike my colleague, I adopt the findings of fact made by the Tax Court of Canada. For these reasons, I would allow the appeal and restore the Tax Court’s judgment.

[145]                     I proceed as follows. First, I reiterate and discuss some of the core principles of the GAAR  jurisprudence, in particular, the principle that the GAAR  cannot be used to override other provisions of the Act . Next, I examine the object, spirit and purpose of s. 111(5) of the Act and describe the problems I see with my colleague’s analysis. Having determined the object, spirit and purpose of the provision, I turn to the question of abuse. First, I lay out the proper standard of review. Then, applying the correct object, spirit and purpose, I conclude that the avoidance transactions at issue do not amount to abusive tax avoidance within the meaning of the GAAR . Finally, I show that even on my colleague’s understanding of the object, spirit and purpose of s. 111(5), a conclusion of abuse is not available without overturning various findings of fact of the Tax Court.

II.            Analysis

A.           The GAAR Cannot Override Parliament’s Intent

[146]                      My colleague’s approach to determining the object, spirit and purpose of s. 111(5) fails to account for one central principle established in Canada Trustco Mortgage Co. v. Canada, 2005 SCC 54, [2005] 2 S.C.R. 601: the GAAR  does not and cannot override Parliament’s specific intent regarding particular provisions of the Act . This principle is integral to the GAAR  jurisprudence and should not be swept aside lightly.

[147]                      It is a longstanding principle of tax law that every person is entitled to order their affairs “so as that the tax attaching under the appropriate Acts is less than it otherwise would be” (Commissioners of Inland Revenue v. Duke of Westminster, [1936] A.C. 1 (H.L.), at p. 19). Of course, this principle is not absolute in the GAAR  context (Lipson, at para. 21; Canada v. Alta Energy Luxembourg S.A.R.L., 2021 SCC 49, at para. 30). The GAAR  is “intended to negate arrangements that would be permissible under a literal interpretation of other provisions of the Income Tax Act , on the basis that they amount to abusive tax avoidance” (Canada Trustco, para. 13). But this purposive flexibility is limited given that courts must “contemporaneously give effect to both the GAAR and the other provisions of the Income Tax Act  relevant to a particular transaction” (para. 13). It is therefore inappropriate to interpret the GAAR  as “implying moral opprobrium regarding the actions of a taxpayer to minimize tax liability utilizing the provisions of the Income Tax Act  in a creative way” (Copthorne Holdings Ltd. v. Canada, 2011 SCC 63, [2011] 3 S.C.R. 721, at para. 65).

[148]                      This is why, since Canada Trustco, this Court has repeatedly affirmed that the object, spirit and purpose analysis is essentially a form of statutory interpretation. There is “but one principle of interpretation: to determine the intent of the legislator” (Canada Trustco, at para. 40). The GAAR  analysis rests on “the same interpretive approach employed by this Court in all questions of statutory interpretation — a ‘unified textual, contextual and purposive approach’” (Copthorne, at para. 70). It “does not rewrite the provisions of the Income Tax Act ; it only requires that a tax benefit be consistent with the object, spirit and purpose of the provisions that are relied upon” (Canada Trustco, at para. 54; Alta Energy, at para. 96). The GAAR  analysis merely seeks to “determine a different aspect of the statute than in other cases” (Copthorne, at para. 70).

[149]                      Since the GAAR analysis is little more than a specialized form of statutory interpretation, the central issue must be determining Parliament’s intent (Canada Trustco, at para. 40). It should not be assumed that the GAAR , as a provision of last resort (Canada Trustco, at para. 21; Copthorne, at para. 66), plays a role in every transaction and in every context. Sometimes, the GAAR  is meant to “address the complexity of transactions which fall outside the scope of specific anti‑avoidance provisions” (Lipson, at para. 47). Other times, Parliament drafts a specific anti-avoidance rule with clear objectives in mind. Applying the GAAR in this latter context would undermine Parliament’s policy choices. Therefore, when it is said that the GAAR  is a provision of last resort, this means only that it is not designed to be Parliament’s first and primary anti‑avoidance measure — nothing more.

[150]                     I agree with my colleague that “there is no bar to applying the GAAR  in situations where the Act  specifies precise conditions that must be met to achieve a particular result, as with a specific anti‑avoidance rule” (para. 71). The GAAR  can indeed supplement a specific anti‑avoidance rule; to say otherwise would “ru[n] contrary to this Court’s jurisprudence” (Rowe J.’s reasons, at para. 72). However, this Court has already acknowledged that a provision’s text can sometimes be conclusive because it is “consistent with and fully explains [the provision’s] underlying rationale” (Copthorne, at para. 110). So while I agree with my colleague’s assertion that the analysis is “case specific” (para. 71), I find it incorrect and unhelpful to definitively and universally declare that technical, carefully drafted provisions “are not immune from abuse” (para. 72), even though Copthorne, which case he cites, specifically acknowledges the very possibility that the underlying rationale of a provision could be no broader than the text itself. That said, I find instead that it is more fruitful to consider when the text is conclusive or not.

[151]                      The key question is whether Parliament specifically intended to prevent or permit a certain type of transaction. Where Parliament has specified precisely what conditions must be satisfied to achieve a particular result, “it is reasonable to assume that Parliament intended that taxpayers would rely on such provisions to achieve the result they prescribe” (Canada Trustco, at para. 11). Although this is not determinative, “where it can be shown that an anti‑avoidance provision has been carefully crafted to include some situations and exclude others, it is reasonable to infer that Parliament chose to limit their scope accordingly” (Minister of National Revenue v. Landrus, 2009 FCA 113, 392 N.R. 54, at para. 47). Therefore, drafting precision is a key factor in revealing whether or not the GAAR  may supplement a specific anti‑avoidance rule. This follows from the core principle that the GAAR  analysis is a species of statutory interpretation: it cannot be used to effectively rewrite a provision.

[152]                     With respect, my colleague applies the GAAR  in a way that not only ignores but also disregards this core principle. He states that the rationale behind a provision must be distinguished from the means used to give effect to that rationale (para. 59). According to him, it is essential to look to the drafting process as it “reflects the task of translating government aims into legislative form in order to create intelligible, legally effective rules” (para. 59). The necessary implication that flows from my colleague’s reasoning is that Parliament is unable to translate its objectives into effective means; the best it can do is choose text that is “most faithful to its objectives” (para. 59 (emphasis added)). In my view, Parliament often intends to do precisely what it actually does. If Parliament crafts a provision carefully, courts should pay close attention to the means chosen and not simply assume that Parliament drafted the provision incompetently.

[153]                     In Alta Energy, this Court affirmed the notion of foreseeability as another important tool to gauge Parliament’s specific intent (para. 82). Sometimes, the absence or presence of a specific anti‑avoidance provision can be “an enlightening contextual and purposive element” that sheds light on Parliament’s intent (para. 82). If “Parliament did not or could not have foreseen the tax strategy employed by the taxpayer”, it is more likely that a gap in the legislation is unintentional, and the GAAR  can be applied to avoidance transactions that exploit that gap (para. 82). The GAAR  was intended to catch unforeseen tax strategies (para. 80) and to allow Parliament to stay one step ahead of ingenious and innovative tax avoidance schemes.

[154]                     But if Parliament drafts a specific anti‑avoidance provision in a way that keeps a highly foreseeable gap open, the gap is more likely to be intentional, and relying on it should not be considered abusive for the purposes of the GAAR . In Alta Energy, this Court declined to use the GAAR  to introduce a “sufficient substantive economic connections” test into the residency requirement of a tax treaty between Canada and Luxembourg. The GAAR  did not apply because “Canada and Luxembourg made a deliberate choice to leave the business property exemption unguarded” in the treaty (para. 82). In other words, the GAAR  cannot be used to punish taxpayers for walking through a door that Parliament intentionally left open.

[155]                     My colleague is essentially applying the same reasoning that this Court expressly rejected in Alta Energy. In that case, my colleague, together with Martin J., dissenting, characterized the relevant articles of the treaty broadly, stating that any other interpretation “renders the GAAR  ‘meaningless’” (para. 133). In their view, “no inference [could] be drawn from the absence of an anti‑avoidance rule designed to counter the particular scheme employed in this case” (para. 150). Thus, “the absence of specific anti‑avoidance rules . . . sheds little light on [the provision’s] underlying rationale” (para. 150). According to the dissent in Alta Energy, drawing such an inference would “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” (para. 150, citing Canada Trustco, at para. 21; Copthorne, at para. 66). In other words, the dissenting view in Alta Energy was that, because the GAAR  is a “provision of last resort”, it has a role to play in each and every case.

[156]                     My colleague follows the same reasoning in this case. He discerns an object, spirit and purpose that conflicts with Parliament’s adoption of a well‑established legal test: de jure control. In his view, Parliament’s chosen control test in s. 111(5)  of the Act  is of little import. Instead, the provision’s “underlying rationale” allows the Minister of National Revenue to apply s. 111(5) in a way that overrides Parliament’s specific intent. Simply stated, this approach is contrary to this Court’s holding in Canada Trustco. It impermissibly elevates the role of the GAAR and enables the Minister to selectively override Parliament. This is not what the GAAR  was intended to do.

B.            The Text, Context and Purpose of Section 111(5)

[157]                     I agree with much of my colleague’s analysis of the text, context and purpose of s. 111(5). However, I disagree with his conclusion that the provision’s rationale is “to deny unused losses to unrelated third parties who take the reins of a corporation and change its business” (para. 118). The analysis below addresses this issue, focusing on the concept of control as this is the main source of disagreement between my colleague and me.

[158]                     I begin my analysis with the observation that s. 111(1) (a) of the Act  allows taxpayers to deduct non‑capital losses for the purpose of computing taxable income for a taxation year. This general rule is restricted by s. 111(5). Upon an acquisition of control, s. 111(5) prevents a corporation from carrying over losses unless the business, carried on by the corporation subject to the change of control, is continued for profit or with a reasonable expectation of profit. Thus, the general loss deductibility provision — s. 111(1)(a) — sets out the default rule subject to the restrictions found in s. 111(5), with exceptions in specific circumstances. Section 111(5) is a specific anti‑avoidance rule that limits what would otherwise be permissible deductions under s. 111(1)(a) (D. G. Duff et al., Canadian Income Tax Law (7th ed. 2023), at pp. 205‑6; T. E. McDonnell, “Legislative Anti‑Avoidance: The Interaction of the New General Rule and Representative Specific Rules”, in Report of Proceedings of the Fortieth Tax Conference (1989), 6:1, at p. 6:18). It bars corporate acquisitions for the singular purpose of accessing tax attributes by restricting the use of those attributes if accessed through the exercise of control. The acquisition of control test and the business continuity requirement are, properly construed, narrow exceptions to the anti‑avoidance provision.

[159]                     It follows that the restrictions on loss carryovers in s. 111(5) are triggered when “control of a corporation has been acquired by a person or group of persons”. “Control” is therefore central to how s. 111(5) functions. The parties in this case disagree as to whether a change of control for the purpose of s. 111(5) is signalled through an acquisition of de jure control or de facto control.

(1)          The Notion of De Jure Control

[160]                     “Control” is not defined in the Act . However, courts have considered this concept on a number of occasions and have determined that “control” for the purposes of the Act  means de jure control. In Buckerfield’s Ltd. v. Minister of National Revenue, [1965] 1 Ex. C.R. 299, the Exchequer Court of Canada held that de jure control refers to the ownership of a sufficient number of shares to have a majority of votes in the election of the corporation’s board of directors.

[161]                     The principle established in Buckerfield’s was affirmed by this Court in Duha Printers (Western) Ltd. v. Canada, [1998] 1 S.C.R. 795. In that case, Iacobucci J. stated that in assessing de jure control, one must consider the corporation’s governing statute, the share register of the corporation and any specific or unique limitation on the majority shareholder’s power to control the election of the board of directors, as manifested in either the corporation’s constating documents or a unanimous shareholder agreement. Iacobucci J. also made it clear that external documents are not relevant to determining de jure control (para. 85). Thus, the voting power attributable to shareholdings, ascertained in light of the corporation’s constating documents and share register, is generally the determinative factor for de jure control.

[162]                     It is useful to recall the rationale for this approach. In Duha Printers, Iacobucci J. distinguished external agreements from a corporation’s constating documents on the basis that the latter create de jure control because they restrain the ability of shareholders to exercise their voting power freely (paras. 65 and 71). He commented at length on the peculiar character of unanimous shareholder agreements, stating that they are “a corporate law hybrid, part contractual and part constitutional in nature” (para. 66). Significantly, Iacobucci J. noted that corporate law offers remedies, such as compliance orders, that effectively enforce rights and obligations under the statutes of a corporation, including a unanimous shareholder agreement. It is important to point out that a compliance order is analogous to a judgment for specific performance (Corporations Act, R.S.M. 1987, c. C225, quoted in Duha Printers, at para. 13; K. P. McGuinness, Canadian Business Corporations Law (3rd ed. 2017), at §22.301).

[163]                     The only exception to the general rule discussed above is found in cases like Minister of National Revenue v. Consolidated Holding Co., [1974] S.C.R. 419, where the capacity of shareholders to act is limited by a trust instrument. In Duha Printers, Iacobucci J. commented on the distinction between a trust instrument and other external documents for the purposes of assessing de jure control, noting that “[a] trust imposes upon the trustee a fiduciary obligation to act within the terms of the trust instrument and for the benefit of the beneficiary” (para. 49). In light of such fiduciary duties, Iacobucci J. concluded that trust instruments limit the capacity of trustees who hold shares to act as “free actors” (para. 49).

[164]                     In contrast to a corporation’s constating documents, external agreements “give rise to obligations that are contractual and not legal or constitutional in nature” (Duha Printers, at para. 59). Consequently, any limitations imposed by such agreements “are limitations freely agreed to by the shareholders, and not at all inconsistent with their de jure power to control the company” (para. 49). Although such agreements do not have an impact on de jure control, they remain relevant in assessing de facto control. This is because control in fact is based on the ability to exercise direct or indirect influence over the corporation’s voting shareholders (Silicon Graphics Ltd. v. Canada, 2002 FCA 260, [2003] 1 F.C. 447, at para. 67). As such, there are a broader range of factors — beyond voting power determined in light of constating documents and the share register — that have the potential to establish de facto control. De facto control thus captures everything that de jure control does not.

(2)          The Binary Nature of Control: De Jure and De Facto Control

[165]                     The legal distinction between constating documents of a corporation and external agreements exposes the binary nature of control, which can be either de jure or de facto. While the former give rise to legal powers over a corporation, the latter do not, such that the distinction between de jure and de facto control goes to the dichotomy between power and influence. The breach of a corporation’s constating documents entitles a complainant to seek a compliance order requiring adherence to the corporate constitution (e.g., Corporations Act, quoted in Duha Printers, at para. 13). In contrast, external agreements do not bind the corporation in the same way: shareholders remain free to exercise de jure control “even if an outside agreement exists to limit actual or de facto control” (Duha Printers, at para. 49). Consequently, the distinction between de jure and de facto control lies in the breadth of factors that can be considered in determining who has control over the corporation.

[166]                     In light of the case law, Parliament has clarified the concept of “control” in order to reach specific legislative goals. Since September 13, 1988, when s. 256(5.1) was introduced, the Act  has been clear as to which provisions refer to the concept of de jure control as opposed to those that involve, rather, the application of de facto control. The use of the phrase “controlled, directly or indirectly in any manner whatever” specifically refers to de facto control, which exists when a party has direct or indirect influence over the corporation which can result in control in fact. This change illustrates the intentionality with which Parliament prescribes control tests, bearing in mind the distinction between de jure and de facto control.

[167]                     The deeming rule found in s. 256(8)  of the Act  is another indication that Parliament maintained and did not abandon the acquisition of  de jure control test in s. 111(5). Rather than replacing the de jure control test, Parliament extended the definition of de jure control to include contingent and future rights to shares under s. 256(8) — but de jure rights nonetheless. Section 256(8) applies when a person acquires certain rights (enumerated in s. 251(5)(b)) in respect of a corporation’s shares and deems the person to be in the same position in relation to the control of the corporation as if the rights had been exercised. Once it is determined that a person has rights under s. 251(5)(b), the de jure control test is applied in relation to those shares. Accordingly, s. 256(8) measures voting power on the basis of shareholdings and captures option agreements that, if exercised, would provide de jure control. It does not capture operational‑level factors that relate to a company’s management.

[168]                      In addition to the above, the advantages of using a bright‑line test are instructive of Parliament’s rationale in selecting a test like de jure control. Clearly defined rules are essential in banking and commercial transactions because they reduce costs related to litigation and promote predictability by allowing all parties to know their rights and obligations in advance. As this Court said in Duha Printers, “a simple test such as that which has been followed since Buckerfield’s is most desirable. If the distinction between de jure and de facto control is to be eliminated at this time, this should be left to Parliament, not to the courts” (para. 52). This is a test that has been recognized and built upon by Parliament (para. 52).

[169]                     The importance of creating transactional certainty and minimizing costs is heightened in the context of a provision that applies exclusively to non‑capital losses. Using a more ambiguous hybrid test would pose commercial risks for parties looking to invest in companies with accrued losses (i.e., risks of reassessment, litigation). In turn, discouraging investments of this type would hinder the rehabilitation of Canadian businesses and undermine economic growth as a whole.

[170]                     In short, Iacobucci J. in Duha Printers took a purposive approach to interpreting the word “control”. Interpreted literally, “control” could mean any type of influence over a corporation (para. 35). Nevertheless, Iacobucci J. concluded that the “general purpose” of s. 111(5) — preventing the transfer of non‑capital losses from one corporation to another — does not expand the meaning of de jure control (para. 88). Therefore, even prior to the adoption of the GAAR , one thing was clear: Parliament intended the acquisition of de jure control — and only the acquisition of de jure control — to trigger the application of s. 111(5). To this day, Parliament continues to rely upon the distinction between de jure and de facto control. And despite my colleague’s suggestion to the contrary, the GAAR  cannot eliminate this distinction.

(3)          Conclusion on the Object, Spirit and Purpose of Section 111(5)

[171]                     In light of the above, the object, spirit and purpose of s. 111(5) is to restrict the use of tax attributes if accessed through an acquisition of de jure control. A textual, contextual and purposive analysis of the relevant provisions reveals that Parliament never intended courts to consider factors other than those related to share ownership.

[172]                     Relying on Alta Energy, my colleague accepts that the GAAR  cannot be used to restrict certain types of tax avoidance strategies if Parliament demonstrates a clear rationale to “permit loss‑trading strategies” or “view[s] [a chosen regime] as an acceptable tradeoff” (para. 119). In his mind, s. 111(5) does not meet these criteria because Parliament “did not make a choice to permit loss‑trading strategies” (para. 119). I strongly disagree. As discussed above, Duha Printers was not based on a literal interpretation of s. 111(5). Understanding the term “control” purposively, this Court found that Parliament specifically intended to allow certain types of legal relationships that create de facto control without resulting in a change of de jure control (paras. 52 and 58). My colleague does not appreciate that Parliament has incorporated a judicially developed meaning of “control” into the provision.

[173]                     My conclusion echoes this Court’s decision in Canada Trustco, which my colleague is unable to distinguish from this case. The key issue in that case was the definition of “cost” for the purpose of s. 20(1)  of the Act . McLachlin C.J. and Major J., writing for the Court, noted that “cost” is a “well‑understood legal concept” that is carefully defined by the Act and the jurisprudence (para. 75). In this context, they held that “nothing in the GAAR  or the object of the [capital cost allowance] provisions . . . permits [this Court] to rewrite them to interpret ‘cost’ to mean ‘amount economically at risk’” (para. 75).

[174]                     The same can be said with respect to the concept of control. Indeed, Parliament’s intent has not changed since Duha Printers. As discussed earlier in these reasons, Parliament has always relied on the de jure and de facto control tests. In 1988, Parliament legislated the concept of de facto control into s. 256(5.1)  of the Act  (Lyrtech RD Inc. v. The Queen, 2014 FCA 267, 2015 D.T.C. 5054, at para. 36). The deeming rule found in s. 256(8) is based and builds upon the notion of de jure control. These provisions show that Parliament has expressly decided to maintain the distinction between de jure and de facto control and continues to rely on it. I do not see how any amendments to the control tests have tarnished the core logic that this Court accepted in Duha Printers. This being so, the GAAR  cannot override Parliament’s intent; it must give effect to it.

[175]                     In sum, a change in control is the singular event that triggers the loss‑trading restrictions found in s. 111(5). The notion of “control” is therefore central to the operation of s. 111(5), and any object, spirit and purpose that omits it lacks coherence. In my opinion, the object, spirit and purpose of s. 111(5) is to deny the use of specified losses of a corporation if a person or group of persons has acquired de jure control of that corporation.

[176]                     My colleague concedes that Parliament has opted for de jure control. But to his mind, “more is needed than the simple fact that Parliament settled on this test to operationalize its intent” (para. 94 (emphasis in original)). He contends that “the de jure control test in s. 111(5) does not, on its own, capture the full range of situations that Parliament sought to target; rather, the test in s. 111(5) is better understood as a rough proxy that seeks to give effect to Parliament’s broader aims” (para. 96 (emphasis in original)). There is only one difficulty with this position, but it is a major one: it completely ignores how Parliament actually defined what constitutes an acquisition of control. With respect, this is wholly inconsistent with this Court’s jurisprudence on interpreting the object, spirit and purpose of a provision under the GAAR .

[177]                     My colleague also tries to downplay the importance of the so‑called “means” of s. 111(5) by omitting the control test from the provision’s rationale. In his view, describing “the object, spirit and purpose of s. 111(5) based on Parliament’s choice of test” would “result in prioritizing the means (the how) over the rationale (the why)” (para. 115 (emphasis in original)). Ironically, his own abuse analysis illustrates how Parliament’s chosen test is central to s. 111(5)’s rationale. While he refuses to formulate the object, spirit and purpose of s. 111(5) as a legal test, my colleague cannot articulate the provision’s rationale without referring to the concept of “acquisition”. What is an “acquisition”? It is merely a synonym for “control”.

[178]                     Despite correctly pointing out that de jure control “remains the standard for the application of s. 111(5)”, my colleague introduces the notion of “functional equivalence” (paras. 117 and 128). This novel concept treats the Investment Agreement as a constating document for the purposes of control (para. 122). The nuance lost on my colleague is that constating documents and external agreements are enforced in radically different ways. That being so, an ordinary contract can never be functionally equivalent to a constating document. In this regard, I see no difference between my colleague’s approach and the Federal Court of Appeal’s “actual control” test. Both ignore the rationale behind the de jure and de facto control tests and erode the distinction between them.

[179]                     The jurisprudence is clear that the GAAR  cannot be invoked to override Parliament’s clear intent. It seems to me, however, that this is what my colleague is doing in this case. Indeed, his approach departs from what is otherwise Parliament’s clear articulation of a de jure control test for restricting losses under s. 111(5). Beyond this, my colleague introduces the novel and unprecedented concept of functional equivalence, which has no known boundaries. With respect, this approach illustrates the potential overriding power of the GAAR , as cautioned against in Canada Trustco.

C.            Abuse Analysis

(1)          Standard of Review

[180]                     My colleague and I agree that whether an avoidance transaction constitutes a misuse or abuse of a provision is a question of mixed fact and law that is reviewable for palpable and overriding error. I cannot agree, however, with his observation that a “trial judge’s factual findings regarding the transactions” may be “more or less relevant when the transactions are considered in light of the correct object, spirit and purpose” (para. 121 (emphasis in original)). With respect, this suggests that it is open to appellate courts to disregard key findings of fact in GAAR  cases.

[181]                     Findings of fact and inferences drawn from those facts can be set aside only if it is established that the trial judge made a palpable and overriding error in reaching those findings and drawing those inferences (Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 S.C.R. 235, at para. 23). Absent a palpable and overriding error, an appellate court cannot reweigh the facts or evidence before the trial court. This is true in any appeal, including one involving the GAAR . I therefore fail to understand why this case in particular warrants the application of a standard less strict than the “palpable and overriding” standard.

[182]                     Whether an avoidance transaction is abusive is a fact‑intensive inquiry that raises a question of mixed fact and law (Canada Trustco, at para. 44). Absent an extricable error of law, the application of the law to the facts is subject to the standard of palpable and overriding error. I note in passing that questions of contractual interpretation are subject to the same deferential standard of review (Sattva Capital Corp. v. Creston Moly Corp., 2014 SCC 53, [2014] 2 S.C.R. 633, at para. 50). As discussed below, it is my view that no such error exists in this case.

(2)          The Avoidance Transactions Do Not Amount to an Abuse of Section 111(5)’s Object, Spirit and Purpose

[183]                      As stated above, the object, spirit and purpose of s. 111(5) is to restrict the use of tax attributes if accessed through an acquisition of de jure control. Since de jure control is an essential element of the object, spirit and purpose of s. 111(5), the key question at this stage is whether Matco acquired de jure control of Deans Knight.

[184]                     Before turning to this question, it should be noted that the relationship between Matco and Deans Knight is the proper focus of the abuse analysis. My colleague seems to imply that the disclosure in the prospectus relating to the foreseeable tax risks associated with the loss transfer is somehow relevant (paras. 21 and 49). However, the Minister did not argue that a change of control occurred as a result of the initial public offering (“IPO”): he argued that a change of control occurred as a result of the avoidance transactions between Matco and Deans Knight. Therefore, given that the prospectus was in connection with the IPO, the disclosure to which my colleague refers is irrelevant here. In any case, disclosures of this kind are commonly included in securities offerings and continuous disclosure documents and are not indicative that a transaction is motivated by abusive tax avoidance.

[185]                     Turning now to the question of abuse, I am of the opinion that the Tax Court’s decision was based on a combination of findings of fact and an interpretation of the Investment Agreement that is supported by the evidence. According to the Tax Court’s factual findings, Matco did not acquire “effective control” of Deans Knight. I see no reviewable error in this conclusion.

[186]                      The Tax Court’s conclusion was largely based on an interpretation of the Investment Agreement and on credibility findings. First, the debenture purchased by Matco under the Investment Agreement was convertible into no more than 35 percent of the voting common shares of Deans Knight (2019 TCC 76, [2019] 4 C.T.C. 2001, at para. 21). At no point did Matco own or have a right to own enough shares to reach a majority shareholder position. Thus, the only key question is whether the evidence supports the contention that Matco had de jure control over the remaining shares of Deans Knight (para. 157).

[187]                      My colleague claims that Matco acquired the “functional equivalent” of de jure control through the Investment Agreement (para. 128). He goes to great lengths to underscore that “[t]he restrictions in favour of Matco resemble the fettering of discretion that would normally occur through a unanimous shareholder agreement and which would lead to an acquisition of de jure control” (para. 132). In his view, those restrictions were so severe that they “effectively neutralized” Deans Knight’s board of directors (para. 132). He therefore concludes that Deans Knight’s status as a “free actor” was “illusory” (para. 134); that its acceptance of the corporate opportunity presented by Matco was a fait accompli (para. 135); that the possibility of Newco receiving a higher price for its shares from a third party was “illusory” (para. 138); and that Newco would receive the same money even if its remaining shares in Deans Knight were not sold shows that the transactions were “designed” to circumvent s. 111(5) (para. 139).

[188]                      In my view, the Investment Agreement is of no relevance to de jure control. In Duha Printers, this Court stated that, contrary to unanimous shareholder agreements, voting agreements give rise to contractual obligations that are neither legal nor constitutional in nature (para. 59). Despite this, my colleague treats the Investment Agreement as sufficient for establishing de jure control. Again, the problem with his conclusion is that external agreements cannot be functionally equivalent to constating documents.

[189]                      But moving past this initial hurdle, my colleague’s approach fails to accord proper deference to the Tax Court’s factual findings. The Tax Court focused on s. 5.8 of the Investment Agreement, which says that “[Newco] shall not be required to sell any of the Remaining Shares prior to or during the Guarantee Period or at any other time” (T.C.C. reasons, at para. 59). The Tax Court found that this term was not a sham or inconsistent with any other rights in the agreement (para. 60). This conclusion was supported by the other terms of the Investment Agreement, which did not give Matco control over the sale of the remaining shares (para. 61) or require that Matco present third parties with a sale opportunity for those shares (para. 62).

[190]                      These findings were supported by oral testimony. David Goold (Deans Knight’s Chief Financial Officer) testified that Newco believed it had a choice whether or not to accept Matco’s offer to purchase the remaining shares (T.C.C. reasons, at paras. 64 and 162‑64). The Tax Court considered the restrictive aspects of the Investment Agreement (paras. 156 and 161) but preferred the testimony of Deans Knight’s witnesses. Put simply, the Tax Court made credibility findings. Credibility findings can be interfered with only if a palpable and overriding error is demonstrated (Housen, at paras. 10‑15 and 19‑25). Since the Tax Court’s factual assessment was not tainted in any way by its characterization of s. 111(5), I see no reason why its findings should be overturned.

[191]                      I see several serious problems with my colleague’s conclusion that the prospect of Newco selling its shares to another party was “illusory” (para. 138). First, my colleague suggests that share ownership is somehow less relevant to the abuse inquiry under s. 111(5). Indeed, he states that “[r]egardless of whether Newco sold its shares to Matco, the actions of the appellant were already locked down by the Investment Agreement, and the core benefits of share ownership (such as the right to dividends) were negated by being subjected to Matco’s approval” (para. 139). This suggests that a shareholder’s right to dividends is somehow relevant to the analysis, but this line of reasoning misses the fact that Parliament’s test for control is squarely focused on voting rights arising from ownership. The “right to dividends” is irrelevant here. Indeed, it is not unusual for a company to issue shares without declaring dividends. The only relevant incidence of ownership is voting power — something that the Investment Agreement did not take away.

[192]                      But putting that aside for a moment, the logical inconsistency in my colleague’s argument shows how voting power is indeed very much relevant to the question on appeal. My colleague finds that the prospect of Newco finding another buyer was “illusory”, but he asserts that his conclusion “is consistent with the circuitous path by which the transactions circumvented s. 111(5)” (para. 139). If the question of whether Newco had the right to retain its shares is irrelevant, why does it matter that this right was “illusory”? Clearly, share ownership is a “decisive” factor in my colleague’s analysis (para. 121).

[193]                      Ultimately and with respect, my colleague’s argument is an attempt to re‑weigh the evidence considered by the Tax Court. The issue of whether Newco retained the right not to sell its remaining shares is a question of mixed fact and law. The Tax Court held that Newco’s right to retain its shares was bona fide and certainly not a sham. Re‑characterizing this right to support the notion that Newco and Deans Knight were intending to abuse s. 111(5) is manifestly counter to the Tax Court’s finding. Considering that my colleague has pointed us to no palpable or overriding error in the Tax Court’s assessment of the evidence, his conclusion requires overturning a key finding of fact made by the Tax Court.

[194]                      My colleague defends his approach by stating that the Tax Court’s finding is tainted by its mischaracterization of the object, spirit and purpose of s. 111(5). He seems to suggest that the Tax Court’s finding of fact can be discounted as it was only relevant to answering the narrow question of whether Matco had acquired control over the majority of the voting shares of the appellant. However, this is simply not the case: the Tax Court “accept[ed] Goold’s testimony that Newco believed it had the choice whether to accept Matco’s offer for the Remaining Shares or not” (para. 164). This is a specific credibility finding on the point that the appellant remained a free actor throughout the transactions. Unlike my colleague, I therefore fail to see how the Tax Court’s mischaracterization of s. 111(5) tainted this important credibility finding.

[195]                      Even if my characterization of s. 111(5)’s object, spirit and purpose is wrong and the object, spirit and purpose of this provision is instead “to prevent corporations from being acquired by unrelated parties in order to deduct their unused losses against income from another business for the benefit of new shareholders” (Rowe J.’s reasons, at para. 113), the appeal should still be allowed.

[196]                      In my view, Matco did not really “acquire” Deans Knight in any practical sense. The Minister’s submissions focused on Matco, and the Tax Court held that Matco was only a facilitator of the transactions (para. 152). The Tax Court found that there was no trade or transfer of non‑capital losses. Matco did not use Deans Knight’s non‑capital losses for its own benefit (para. 152). The post‑IPO shareholders were the only ones that directly received the benefits of ownership. At best, one could argue that the Investment Agreement “locked down” Deans Knight so that Matco could market it for a subsequent acquisition. However, it requires a great leap in logic to infer that Matco “acquired” Deans Knight. Such a leap would create uncertainty as to what exactly constitutes abuse: at what point does “tak[ing] the reins” of a corporation result in the acquisition of that corporation (Rowe J.’s reasons, at para. 118)? Under my colleague’s approach, the existence of abusive tax avoidance is, at best, unclear. In such circumstances, “the benefit of the doubt goes to the taxpayer” (Canada Trustco, at para. 66).

[197]                      For these reasons, I am of the view that the avoidance transactions did not frustrate the rationale of s. 111(5). There was therefore no abuse, and consequently, I would allow the appeal and restore the judgment of the Tax Court of Canada.

 

APPENDIX

Income Tax Act ,  R.S.C. 1985, c. 1 (5th Supp .)

 

2 (1) An income tax shall be paid, as required by this Act , on the taxable income for each taxation year of every person resident in Canada at any time in the year.

(2) The taxable income of a taxpayer for a taxation year is the taxpayer’s income for the year plus the additions and minus the deductions permitted by Division C.

(3) Where a person who is not taxable under subsection 2(1) for a taxation year

(a) was employed in Canada,

(b) carried on a business in Canada, or

(c) disposed of a taxable Canadian property,

at any time in the year or a previous year, an income tax shall be paid, as required by this Act , on the person’s taxable income earned in Canada for the year determined in accordance with Division D.

 

3 The income of a taxpayer for a taxation year for the purposes of this Part is the taxpayer’s income for the year determined by the following rules:

(a) determine the total of all amounts each of which is the taxpayer’s income for the year (other than a taxable capital gain from the disposition of a property) from a source inside or outside Canada, including, without restricting the generality of the foregoing, the taxpayer’s income for the year from each office, employment, business and property,

(b) determine the amount, if any, by which

(i) the total of

(A) all of the taxpayer’s taxable capital gains for the year from dispositions of property other than listed personal property, and

(B) the taxpayer’s taxable net gain for the year from dispositions of listed personal property,

exceeds

(ii) the amount, if any, by which the taxpayer’s allowable capital losses for the year from dispositions of property other than listed personal property exceed the taxpayer’s allowable business investment losses for the year,

(c) determine the amount, if any, by which the total determined under paragraph (a) plus the amount determined under paragraph (b) exceeds the total of the deductions permitted by Subdivision E in computing the taxpayer’s income for the year (except to the extent that those deductions, if any, have been taken into account in determining the total referred to in paragraph (a), and

(d) determine the amount, if any, by which the amount determined under paragraph (c) exceeds the total of all amounts each of which is the taxpayer’s loss for the year from an office, employment, business or property or the taxpayer’s allowable business investment loss for the year,

and for the purposes of this Part,

(e) where an amount is determined under paragraph (d) for the year in respect of the taxpayer, the taxpayer’s income for the year is the amount so determined, and

(f) in any other case, the taxpayer shall be deemed to have income for the year in an amount equal to zero.

 

111 (1) For the purpose of computing the taxable income of a taxpayer for a taxation year, there may be deducted such portion as the taxpayer may claim of the taxpayer’s

(a) non-capital losses for the 20 taxation years immediately preceding and the 3 taxation years immediately following the year;

. . .

(5) Where, at any time, control of a corporation has been acquired by a person or group of persons, no amount in respect of its non-capital loss or farm loss for a taxation year ending before that time is deductible by the corporation for a taxation year ending after that time and no amount in respect of its non-capital loss or farm loss for a taxation year ending after that time is deductible by the corporation for a taxation year ending before that time except that

(a) such portion of the corporation’s non-capital loss or farm loss, as the case may be, for a taxation year ending before that time as may reasonably be regarded as its loss from carrying on a business and, where a business was carried on by the corporation in that year, such portion of the non-capital loss as may reasonably be regarded as being in respect of an amount deductible under paragraph 110(1)(k) in computing its taxable income for the year is deductible by the corporation for a particular taxation year ending after that time

(i) only if that business was carried on by the corporation for profit or with a reasonable expectation of profit throughout the particular year, and

(ii) only to the extent of the total of the corporation’s income for the particular year from that business and, where properties were sold, leased, rented or developed or services rendered in the course of carrying on that business before that time, from any other business substantially all the income of which was derived from the sale, leasing, rental or development, as the case may be, of similar properties or the rendering of similar services; and

(b) such portion of the corporation’s non-capital loss or farm loss, as the case may be, for a taxation year ending after that time as may reasonably be regarded as its loss from carrying on a business and, where a business was carried on by the corporation in that year, such portion of the non-capital loss as may reasonably be regarded as being in respect of an amount deductible under paragraph 110(1)(k) in computing its taxable income for the year is deductible by the corporation for a particular year ending before that time

(i) only if throughout the taxation year and in the particular year that business was carried on by the corporation for profit or with a reasonable expectation of profit, and

(ii) only to the extent of the corporation’s income for the particular year from that business and, where properties were sold, leased, rented or developed or services rendered in the course of carrying on that business before that time, from any other business substantially all the income of which was derived from the sale, leasing, rental or development, as the case may be, of similar properties or the rendering of similar services.

 

245 (1) In this section,

tax benefit means a reduction, avoidance or deferral of tax or other amount payable under this Act  or an increase in a refund of tax or other amount under this Act , and includes a reduction, avoidance or deferral of tax or other amount that would be payable under this Act  but for a tax treaty or an increase in a refund of tax or other amount under this Act  as a result of a tax treaty; (avantage fiscal)

tax consequences to a person means the amount of income, taxable income, or taxable income earned in Canada of, tax or other amount payable by or refundable to the person under this Act , or any other amount that is relevant for the purposes of computing that amount; (attribut fiscal)

transaction includes an arrangement or event. (opération)

(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.

(3) An avoidance transaction means any transaction

(a) that, but for this section, would result, directly or indirectly, in a tax benefit, unless the transaction may reasonably be considered to have been undertaken or arranged primarily for bona fide purposes other than to obtain the tax benefit; or

(b) that is part of a series of transactions, which series, but for this section, would result, directly or indirectly, in a tax benefit, unless the transaction may reasonably be considered to have been undertaken or arranged primarily for bona fide purposes other than to obtain the tax benefit.

(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

(i) this Act ,

(ii) the Income Tax Regulations,

(iii) the Income Tax Application Rules,

(iv) a tax treaty, or

(v) any other enactment that is relevant in computing tax or any other amount payable by or refundable to a person under this Act  or in determining any amount that is relevant for the purposes of that computation; or

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

(5) Without restricting the generality of subsection (2), and notwithstanding any other enactment,

(a) any deduction, exemption or exclusion in computing income, taxable income, taxable income earned in Canada or tax payable or any part thereof may be allowed or disallowed in whole or in part,

(b) any such deduction, exemption or exclusion, any income, loss or other amount or part thereof may be allocated to any person,

(c) the nature of any payment or other amount may be recharacterized, and

(d) the tax effects that would otherwise result from the application of other provisions of this Act  may be ignored,

in determining the tax consequences to a person as is reasonable in the circumstances in order to deny a tax benefit that would, but for this section, result, directly or indirectly, from an avoidance transaction.

 

[249] (4) Where at any time control of a corporation . . . is acquired by a person or group of persons, for the purposes of this Act ,

(a) subject to paragraph 249(4)(c), the taxation year of the corporation that would, but for this paragraph, have included that time shall be deemed to have ended immediately before that time;

(b) a new taxation year of the corporation shall be deemed to have commenced at that time;

(c) subject to paragraph 128(1)(d), section 128.1, and paragraphs 142.6(1)(a) and 149(10)(a), and notwithstanding subsections 249(1) and 249(3), where the taxation year of the corporation that would, but for this subsection, have been its last taxation year that ended before that time would, but for this paragraph, have ended within the 7-day period that ended immediately before that time, that taxation year shall, except where control of the corporation was acquired by a person or group of persons within that period, be deemed to end immediately before that time where the corporation so elects in its return of income under Part I for that taxation year; and

(d) for the purpose of determining the corporation’s fiscal period after that time, the corporation shall be deemed not to have established a fiscal period before that time.

 

[251] (5) For the purposes of subsection 251(2) and the definition Canadian-controlled private corporation in subsection 125(7),

. . .

(b) where at any time a person has a right under a contract, in equity or otherwise, either immediately or in the future and either absolutely or contingently,

(i) to, or to acquire, shares of the capital stock of a corporation or to control the voting rights of such shares, the person shall, except where the right is not exercisable at that time because the exercise thereof is contingent on the death, bankruptcy or permanent disability of an individual, be deemed to have the same position in relation to the control of the corporation as if the person owned the shares at that time,

(ii) to cause a corporation to redeem, acquire or cancel any shares of its capital stock owned by other shareholders of the corporation, the person shall, except where the right is not exercisable at that time because the exercise thereof is contingent on the death, bankruptcy or permanent disability of an individual, be deemed to have the same position in relation to the control of the corporation as if the shares were so redeemed, acquired or cancelled by the corporation at that time;

(iii) to, or to acquire or control, voting rights in respect of shares of the capital stock of a corporation, the person is, except where the right is not exercisable at that time because its exercise is contingent on the death, bankruptcy or permanent disability of an individual, deemed to have the same position in relation to the control of the corporation as if the person could exercise the voting rights at that time, or

(iv) to cause the reduction of voting rights in respect of shares, owned by other shareholders, of the capital stock of a corporation, the person is, except where the right is not exercisable at that time because its exercise is contingent on the death, bankruptcy or permanent disability of an individual, deemed to have the same position in relation to the control of the corporation as if the voting rights were so reduced at that time; . . . .

 

[256] (5.1) For the purposes of this Act , where the expression “controlled, directly or indirectly in any manner whatever,” is used, a corporation shall be considered to be so controlled by another corporation, person or group of persons (in this subsection referred to as the “controller”) at any time where, at that time, the controller has any direct or indirect influence that, if exercised, would result in control in fact of the corporation, except that, where the corporation and the controller are dealing with each other at arm’s length and the influence is derived from a franchise, licence, lease, distribution, supply or management agreement or other similar agreement or arrangement, the main purpose of which is to govern the relationship between the corporation and the controller regarding the manner in which a business carried on by the corporation is to be conducted, the corporation shall not be considered to be controlled, directly or indirectly in any manner whatever, by the controller by reason only of that agreement or arrangement.

 

(7) For the purposes of . . . sectio[n] 111 and . . . this subsection,

(a) control of a particular corporation shall be deemed not to have been acquired solely because of

(i) the acquisition at any time of shares of any corporation by

(A) a particular person who acquired the shares from a person to whom the particular person was related (otherwise than because of a right referred to in paragraph 251(5)(b)) immediately before that time,

(B) a particular person who was related to the particular corporation (otherwise than because of a right referred to in paragraph 251(5)(b)) immediately before that time,

(C) an estate that acquired the shares because of the death of a person, or

(D) a particular person who acquired the shares from an estate that arose on the death of another person to whom the particular person was related, or

. . .

(b) where at any time 2 or more corporations (each of which is referred to in this paragraph as a “predecessor corporation”) have amalgamated to form one corporate entity (in this paragraph referred to as the “new corporation”),

. . .

(ii) a person or group of persons that controls the new corporation immediately after the amalgamation and did not control a predecessor corporation immediately before the amalgamation is deemed to have acquired immediately before the amalgamation control of the predecessor corporation and of each corporation it controlled immediately before the amalgamation (unless the person or group of persons would not have acquired control of the predecessor corporation if the person or group of persons had acquired all the shares of the predecessor corporation immediately before the amalgamation), and

(iii) control of a predecessor corporation and of each corporation it controlled immediately before the amalgamation is deemed to have been acquired immediately before the amalgamation by a person or group of persons

(A) unless the predecessor corporation was related (otherwise than because of a right referred to in paragraph 251(5)(b)) immediately before the amalgamation to each other predecessor corporation,

. . .

(8) Where at any time a taxpayer acquires a right referred to in paragraph 251(5)(b) in respect of a share and it can reasonably be concluded that one of the main purposes of the acquisition is

(a) to avoid any limitation on the deductibility of any non-capital loss, net capital loss, farm loss or any expense or other amount referred to in subsection 66(11), 66.5(3) or 66.7(10) or 66.7(11),

. . .

the taxpayer is deemed to be in the same position in relation to the control of the corporation as if the right were immediate and absolute and as if the taxpayer had exercised the right at that time for the purpose of determining whether control of a corporation has been acquired for the purposes of . . . sectio[n] 111 . . . .

                    Appeal dismissed with costs, Côté J. dissenting.

                    Solicitors for the appellant: Burnet, Duckworth & Palmer, Calgary.

                    Solicitor for the respondent: Attorney General of Canada — Department of Justice, British Columbia Regional Office, Vancouver.

                    Solicitor for the intervener the Attorney General of Ontario: Ministry of the Attorney General — Crown Law Office — Civil, Toronto.

                    Solicitors for the intervener the Canadian Chamber of Commerce: Borden Ladner Gervais, Toronto.

                    Solicitors for the intervener the Tax Executives Institute, Inc.: Osler, Hoskin & Harcourt, Toronto.

                    Solicitor for the intervener Agence du Revenu du Québec: Agence du Revenu du Québec, Direction du contentieux fiscal et civil, Montréal.