Gladwin Realty Corporation v. Canada

Gladwin Realty Corporation v. Canada
Court (s) Database
Federal Court of Appeal Decisions
Date
2020-09-14
Neutral citation
2020 FCA 142
File numbers
A-138-19
Decision Content
Date: 20200914
Docket: A-138-19
Citation: 2020 FCA 142
CORAM:
NOËL C.J.
DE MONTIGNY J.A.
LEBLANC J.A.
BETWEEN:
THE GLADWIN REALTY CORPORATION
Appellant
and
HER MAJESTY THE QUEEN
Respondent
Heard at Ottawa, Ontario, on June 8, 2020.
Judgment delivered at Ottawa, Ontario, on September 14, 2020.
REASONS FOR JUDGMENT BY:
NOËL C.J.
CONCURRED IN BY:
DE MONTIGNY J.A.
LEBLANC J.A.
Date: 20200914
Docket: A-138-19
Citation: 2020 FCA 142
CORAM:
NOËL C.J.
DE MONTIGNY J.A.
LEBLANC J.A.
BETWEEN:
THE GLADWIN REALTY CORPORATION
Appellant
and
HER MAJESTY THE QUEEN
Respondent
REASONS FOR JUDGMENT
NOËL C.J.
[1] This is an appeal brought by The Gladwin Realty Corporation Inc. (Gladwin or the appellant) from a decision of the Tax Court of Canada (2019 TCC 62) wherein Hogan J. (the Tax Court judge) dismissed the appellant’s appeal from a Notice of determination issued by the Minister of National Revenue (the Minister) with respect to its 2008 taxation year. The Notice of determination was issued pursuant to subsection 152(1.11) and the General Anti-Avoidance Rule (GAAR) found in section 245 of the Income Tax Act, R.C.S. 1985, c. 1 (5th Supp.) (the Act).
[2] The Tax Court judge held that the avoidance transactions undertaken by Gladwin, which involved rolling real estate assets into a partnership, disposing of these assets, distributing the proceeds to the limited partner, triggering the application of subsection 40(3.1) and making the election under subsection 40(3.12), to be abusive. The order and timing of the series of transactions allowed the appellant to distribute the entire capital gain realized from the sale to its corporate shareholder as a tax-free capital dividend.
[3] The appellant contends that the avoidance transactions are not abusive because they produce results mandated under the Act. It argues that the deemed capital gain arising from the application of subsection 40(3.1) is not artificial and that the Capital Dividend Account (CDA) mechanism allows taxpayers to benefit from the timing of the realization of capital gains and losses.
[4] The Crown argues that the appellant deliberately triggered the deemed capital gain provided for under subsection 40(3.1), thereby artificially inflating the appellant’s CDA so that the entire capital gain realized for the sale of the property could be paid out tax-free in the form of a capital dividend. This artificial inflation led to the abuse of the CDA regime as well as the negative adjusted cost base (ACB) rules set out in the Act.
[5] For the reasons which follow, I propose that the appeal be dismissed.
[6] The provisions of the Act that are relevant to the analysis appear at the end of the reasons.
FACTS
[7] The factual background is set out in a Partial Agreed Statement of Facts (Appeal Book, Vol. 1, pp. 98-109). Of significance for present purposes is the following series of transactions which were found to have been entered into “with clockwork precision” with the view of paying no tax on a capital gain of $23,346,822 resulting from the sale of the real estate assets acquired six years earlier (Reasons, para. 23). This objective was achieved by the planned inflation of the appellant’s CDA.
[8] The first step of the plan was the incorporation of Gladwin GP Inc. (Gladwin GP) on February 6, 2007, and its utilization as the general partner of the partnership which, in turn, was created on March 1, 2007. The appellant was a limited partner holding a 99% interest in the partnership, and Gladwin GP held the remaining 1% as general partner. The Tax Court judge’s finding that the partnership was created for the purpose of securing the tax benefit and that its role was transitory in nature is uncontested (Reasons, para. 25).
[9] On April 10, 2007, the real estate assets were transferred to the newly formed partnership on a tax-free basis by way of a rollover pursuant to subsection 97(2) of the Act, thereby ensuring that the gain embedded in the real estate assets would be realized by the partnership, rather than by the appellant directly. The partnership sold the real estate assets to an arm’s length party on August 8, 2007, thereby realizing a capital gain of $23,346,822.
[10] On that same day, the partnership lent an amount of $24,463,142 to Shabholdings Inc. (Shabholdings), the appellant’s direct shareholder, in return for a promissory note.
[11] On September 26, 2007, the appellant was continued under the British Virgin Island legislation and, as a result, gave up its Canadian-controlled private corporation status. In so doing, the appellant avoided the additional tax on investment income (including taxable capital gains) under section 123.3 of the Act.
[12] On September 28, 2007—during its first fiscal period—the partnership distributed $24,647,031 to the appellant. This distribution, which was partially satisfied by the assignment of the promissory note issued by Shabholdings, caused the ACB of the appellant's interest in the partnership to become negative by an amount of $24,311,654, pursuant to paragraph 53(2)(c) of the Act.
[13] As a result on October 1, 2007, the date that coincides with the end of the partnership’s first fiscal period, the appellant realized a deemed capital gain in the amount equal to the negative ACB of its interest in the partnership pursuant to subsection 40(3.1). Half of this amount (i.e. $12,155,827) was added to the appellant’s CDA in conformity with subsection 89(1).
[14] On the same day, the partnership allocated the capital gain of $23,346,822 realized earlier on the disposition of the real estate assets to the appellant. Here again, half of the gain was added to the appellant’s CDA (i.e. $11,673,410), thereby increasing its CDA balance to $23,829,237.
[15] On May 30, 2008, the appellant paid a capital dividend to Shabholdings equal to its CDA balance thereby reducing its CDA to nil. Payment was effected by distributing a portion of the promissory note. On the same day, Shabholdings elected that capital dividends in the same amount be paid to its shareholders—all corporations.
[16] For its taxation year ending on September 30, 2008—during the partnership’s second fiscal period—the appellant elected to realize a capital loss in the amount of $24,311,653, pursuant to subsection 40(3.12), an amount equal to the capital gain it triggered pursuant to subsection 40(3.1) earlier in that taxation year. As a result, the ACB of the appellant's interest was decreased by a corresponding amount by virtue of subparagraph 53(2)(c)(i.2). This election had the effect of bringing the appellant’s CDA balance to a negative amount of roughly $12,000,000.
[17] Both the appellant and the partnership ceased operations after the transactions in issue were completed and have since remained inactive (Partial Agreed Statement of Facts, Appeal Book, Vol. 1, para. gg, p. 105).
[18] In the end, the appellant’s CDA was increased twice by roughly the same amount, with the result that an amount equal to the entire capital gain arising from the disposition of the real estate assets was distributed on a tax-free basis.
[19] By Notice of determination issued on March 27, 2014, the appellant’s CDA available for distribution during its 2008 taxation year was reduced by $12,153,827, thereby eliminating the increase resulting from the triggered application of subsections 40(3.1).
[20] The appeal before the Tax Court ensued.
DECISION UNDER APPEAL
[21] At the beginning of his reasons, the Tax Court judge noted, in a footnote, that the tax benefit in this case was the avoidance of the additional tax on excessive elections provided for under subsection 184(2) on the payment of the excessive capital dividend to the appellant (Reasons, para. 6). Given the appellant’s concession that the series of transactions were avoidance transactions within the meaning of subsection 245(3), the only issue to be addressed was whether those transactions were abusive (Reasons, para. 32).
[22] He began by conducting an object, spirit and purpose analysis of the relevant provisions in order to identify their underlying rationale. Dealing with the CDA regime, the Tax Court judge explained that the purpose of the CDA, as defined under subsection 89(1), was to allow “private corporations to keep track of certain types of tax-free surpluses accumulated over time […] [and] to determine its CDA balance at a particular time so that it may elect in a prescribed form to pay a tax-free capital dividend to its shareholders without incurring a liability under Part III of the Act” (Reasons, para. 39).
[23] He went on to explain that the legislative history of the CDA regime confirmed that subsections 89(1), 83(2), and 184(2) were implemented to give effect to what is commonly referred to as the principle of integration. Indeed, those provisions were enacted as a result of the 1966 Report of the Royal Commission on Taxation, which among other things recommended that income be taxed at the same rate whether earned directly by an individual, or indirectly through a corporation. Consistent with the principle of integration, the CDA regime was put in place to ensure that the tax-free portion of capital gain could be distributed to an individual shareholder without any tax (Reasons, paras. 41-42). Although integration is not perfectly achieved in all circumstances, the Tax Court judge held that the CDA provisions were nevertheless intended to promote integration (Reasons, para. 46).
[24] The Tax Court judge concluded his analysis of the CDA regime by considering the subsequent legislative amendment brought to subsection 89(1) in 2013 and held that they were not instructive in determining whether the avoidance transactions were abusive (Reasons, para. 47).
[25] The Tax Court judge then considered the effect of subsection 40(3.1) as well as the 1994 Budget Supplementary Information that accompanied the enactment of this provision (1994 Budget Supplementary Information, Department of Finance, Tax Measures, February 22, 1994, p. 42). He observed that subsection 40(3.1) had been enacted to put an end to tax shelter planning arrangements that took advantage of the fact that partnership interests were excluded from the application of subsection 40(3), a provision which deems a capital gain to be realized whenever the ACB of a capital property of a taxpayer becomes negative (Reasons, paras. 49-51).
[26] He further explained that subsection 40(3.1) applies to limited partners or other passive partners in circumstances where the ACB of their partnership interest is negative at the end of the partnership’s fiscal period by deeming them to realize a capital gain equal to this negative ACB (Reasons, para. 55). Relying on the 1994 Budget Supplementary Information, the Tax Court judge concluded that the “purpose and effect of subsection 40(3.1) are to dissuade taxpayers from extracting from a partnership on a tax-free basis funds in excess of their investment in the partnership” (Reasons, para. 58).
[27] Turning to subsection 40(3.12), he found, based on the 1994 Budget Supplementary Information and the harsh effect that subsection 40(3.1) can have, that this provision operates as a relieving provision by allowing taxpayers to offset the deemed gain realized pursuant to subsection 40(3.1) and recover the tax paid on the gain (Reasons, paras. 63-64).
[28] The Tax Court judge then turned to the second part of the abuse analysis, namely whether the appellant frustrated the underlying rationale of the CDA regime by the triggering of the second capital gain and by electing to offset it by the deemed capital loss under subsections 40(3.1) and 40(3.12) (Reasons, para. 68).
[29] The Tax Court judge answered this question in the affirmative based on his finding that the avoidance transactions were specifically designed to achieve a result that was inconsistent with the rationale underlying subsections 40(3.1) and 40(3.12) and the CDA regime. He explained that, but for the application of the GAAR, the appellant would be able to make a tax-free distribution equal to the entire capital gain realized from the sale of the real estate assets (Reasons, para. 85-86).
[30] Finally, the Tax Court judge concluded that subsections 40(3.1) and 40(3.12) were not intended to allow the appellant to achieve the tax benefit obtained. According to him, the result achieved was inconsistent with the underlying rationale of these provisions as well as the underlying rationale of the CDA regime (Reasons, paras. 88-89).
POSITION OF THE APPELLANT
[31] In support of its appeal, the appellant submits that the Tax Court judge erred in concluding that the series of transactions was abusive as he failed to consider the existence of the two different assets, i.e. the real estate assets and the appellant’s partnership interest which generated distinct gains (Memorandum of the Appellant, para. 35). Although the Tax Court judge held with apparent conviction that the series of transactions was abusive, the reasons offered in support of this conclusion do not follow the principled approach that is to be used in conducting the abuse analysis.
[32] When regard is had to the GAAR framework, the appellant argues that taxpayers are allowed to carry out transactions that will minimize their tax liability, and that “the term abuse does not imply a moral opprobrium regarding the actions of the taxpayer in doing so.” It insists that determining the underlying rationale of the provisions at issue should be an objective exercise rather than a “value judgment of what is right or wrong” (Memorandum of the Appellant, paras. 38-40). As a last observation on the GAAR framework, the appellant notes that section 245 is a measure of last resort, and that in the case at hand, the appropriate remedy was a legislative amendment in order to change the existing policy, a step which took place when Parliament altered the definition of CDA in subsection 89(1) in 2013 (Memorandum of the Appellant, para. 44).
[33] The appellant does not take issue with the inner workings of the CDA regime as described by the Tax Court judge. It argues that the definition of CDA is nothing more than a formula provided to compute the CDA balance at a point in time. Further, it contends that this regime is not concerned with the origins of the capital gains and losses, but only by the maximum amount a corporation can pay out tax-free to a shareholder. There is therefore no distinction to be made between gains arising from a true disposition and a deemed gain, nor is it relevant to consider whether Part I tax has been paid on the taxable portion of the gain. Finally, it submits that section 184 also is of no relevance in determining what should be included in the CDA balance, as it is simply the remedy chosen by Parliament to address excessive payments (Memorandum of the Appellant, paras. 55-61).
[34] Turning to subsections 40(3.1) and 40(3.12) (also referred to as the negative ACB rules), the appellant acknowledges that subsection 40(3.1) is a specific anti-avoidance rule and goes on to explain its application as well as its interaction with subsection 40(3.12) the same way as does the Tax Court judge. The appellant, however, relies on the last quoted paragraph of 1994 Budget Supplementary Information and insists that the mischief that the negative ACB rules sought to address was the circumvention of the partnership “at-risk” rules by “having partnership loss allocations precede distributions” (Memorandum of the Appellant, para. 68). It points out that in the instant case, the distribution preceded the allocation and that, as a specific anti-avoidance provision, subsection 40(3.1) operated exactly as it should. Regarding subsection 40(3.12), the appellant posits that the loss it elected pursuant to that provision did trigger a reduction of its CDA pursuant to subsection 89(1) (Memorandum of the Appellant, paras. 69-71).
[35] When regard is had to the Tax Court judge’s view that the series of transactions had enabled the appellant to distribute the entire capital gain resulting from the sale of the real estate assets without tax, the appellant argues that he is ignoring bona fide legal relationships, contrary to Shell Canada Ltd. v. Canada, [1999] 3 S.C.R. 622 (Memorandum of the Appellant, para. 74). Given that the two realized capital gains reflected “economic realities” (i.e. the one resulting from the sale of the real estate assets and the deemed gain resulting from the negative ACB), the appellant argues that both these gains must be recognized in computing the CDA balance (Memorandum of the Appellant, para. 76). Additionally, the appellant submits that a failure to take into account both gains ignores two further Supreme Court decisions holding that separate transactions must be viewed independently (Friedberg v. Canada, [1993] 4 S.C.R. 285 and Singleton v. the Queen, 2001 SCC 61, [2001] 2 S.C.R. 1046) (Memorandum of the Appellant, paras. 76-81). As well, the appellant relies on Canada Trustco Mortgage Co. v. The Queen, 2005 SCC 54, [2005] 2 S.C.R. 601 [Canada Trustco] to argue that the economic substance of a transaction should not be “viewed in isolation from a textual, contextual and purposive interpretation” of the relevant provisions (Memorandum of the Appellant, para. 87 citing Canada Trustco at para. 76).
[36] Lastly, the appellant contends that the Tax Court judge erred in holding that the series of transaction had led to “over-integration” and that, in any case, integration is a mere concept as there is no broad or overarching integration policy in the Act (Memorandum of the Appellant, paras. 89-92). If such a policy ever existed, submits the appellant, the question is whether timing the CDA is inappropriate, specifically, whether paying a dividend when the underlying tax has not been paid frustrates this policy (Memorandum of the Appellant, paras. 97-98). Pointing to specific provisions of the Act that promote integration, the appellant posits that “Parliament chose not to adopt a general integration regime, and as such, there can be no general integration policy in the Act” (Memorandum of the Appellant, para. 108).
[37] As to the 2013 legislative amendment that was brought to the CDA definition set out in subsection 89(1), the appellant argues that the Tax Court judge erred in holding that this amendment was not instructive as it “clearly modified the existing law” (Memorandum of the Appellant, para. 110). The appellant submits that reading the new enactment as clarifying the existing law amounts to using this amendment in order to support a finding of abuse, a proposition that runs against the decision of this Court in Univar Holdco Canada ULC v. Canada, 2017 FCA 207, [2019] 2 F.C.R. 569 [Univar] (Memorandum of the Appellant, para. 128 citing Univar at para. 29).
POSITION OF THE CROWN
[38] The Crown begins by citing Lipson v. Canada, 2009 SCC 1, [2009] 1 S.C.R. 3 [Lipson] at paras. 42 and 47 and the decision of this Court in Fiducie financière Satoma v. Canada, 2018 FCA 74 [Satoma Trust] at para. 52 for the proposition that using an anti-avoidance provision in order to obtain a tax benefit gives rise to an abuse. The Crown adds that subsection 40(3.1) is an anti-avoidance provision that was used for that purpose (Memorandum of the Crown, Overview, p. 1).
[39] The Crown further contends that the GAAR framework requires going behind the words of the provisions relied on by the taxpayer, in order to determine their underlying rationale. The Crown argues that the “Tax Court judge correctly concluded that the underlying rationale of the CDA regime is to promote integration and not over-integration” (Memorandum of the Crown, para. 41). In that respect, the Crown submits that the “CDA regime, which includes subsections 83(2), 89(1) and 184(2), is an integral part of the Act’s mechanism with respect to integration” and that these provisions viewed as a whole allow the non-taxable portion of capital gains realized by a private corporation to flow through an individual taxpayer without attracting further tax (Memorandum of the Crown, para. 54).
[40] With respect to the rationale underlying the CDA regime, the Crown submits that the intention was to provide a mechanism ensuring that the non-taxable portion of capital gains retains its character in the hands of the recipient, and that only this non-taxable half is capable of being distributed tax-free (Memorandum of the Crown, para. 60).
[41] As to the legislative amendment brought to subsection 89(1), the Crown submits that the amended text expressly provides for a rationale already found in the Act (Memorandum of the Crown, para. 65).
[42] With respect to the negative ACB rules, the Crown agrees with the Tax Court judge’s determination of the rationale underlying subsections 40(3.1) and 40(3.12), namely to prevent passive investors to receive tax-free distributions in excess of their investment and the undistributed income allocated to them (Memorandum of the Crown, paras. 66, 81-82).
[43] Further, the Crown submits that the Tax Court judge did not err when he concluded that the overall result of the series of transactions frustrated the underlying rationale of both the CDA regime and the negative ACB rules (Memorandum of the Crown, para. 84). Specifically, the Crown argues that the abuse resides in the fact that the appellant triggered subsection 40(3.1), an anti-avoidance provision, which, combined with the election provided for under subsection 40(3.12), offends the purpose of both these provisions (Memorandum of the Crown, para. 90).
[44] Finally, the Crown argues that the Tax Court judge did not reintroduce the doctrine of economic reality in the GAAR analysis, but rather properly conducted the mandated analysis (Memorandum of the Crown, para. 96).
ANALYSIS
[45] In order to conclude that GAAR applies, the Tax Court judge had to answer three questions in the affirmative: was there a tax benefit? If so, were the transactions which gave rise to this benefit avoidance transactions? If so, were the avoidance transactions abusive? (Copthorne Holdings Ltd. v. Canada, 2011 SCC 63, [2011] 3 S.C.R. 721 [Copthorne], para. 33, citing Canada Trustco at paras. 18, 21, 36).
A. The Tax Benefit
[46] As to the first two questions, there is agreement that we are dealing with avoidance transactions that resulted in a tax benefit, but there is no agreement as to what this benefit actually is. The Tax Court judge surmised that if the appeal fails, the Minister will assess an additional tax on excessive elections under subsection 184(2) unless the appellant elects to treat the excessive capital dividend as a taxable dividend. He concluded from this that the tax benefit was the avoidance of the additional tax on the excessive amount (Reasons, para. 6). The Crown agrees with this characterization of the tax benefit (Memorandum of the Crown, para. 26).
[47] The appellant for its part contends that the tax benefit results from the CDA increase in the amount of $12,155,827 (Memorandum of the Appellant, para. 1). However, it is now established that the modification of tax attributes, such as an increase in a taxpayer’s CDA, does not give rise to a tax benefit unless and until a capital dividend is paid out of that account to a recipient capable of benefiting from its tax-free character (see 1245959 Alberta Ltd. v. Canada (Attorney General), 2018 FCA 114 [Perry Wild]). Although distributions have been made from the appellant’s CDA, the recipients so far are all related corporations which can otherwise receive inter-corporate dividends on a tax-free basis (subsection 112(1)).
[48] The tax benefit identified by the Tax Court judge, and agreed to by the Crown, no more qualifies as a tax benefit. The GAAR may only be used to deny a tax benefit that results from an avoidance transaction. The tax benefit in the form of the avoidance of the additional tax on excessive elections provided for under subsection 184(2) would result from the Notice of determination itself which reduced the appellant’s CDA by half of the second capital gain in order to deny the tax benefit. A tax benefit cannot be the result of a Notice of determination issued under the GAAR as the purpose of such a determination is to deny the tax benefit, not to bring the benefit into existence.
[49] Faced with the parties’ desire to obtain a resolution in this matter and the absence of a tax benefit, counsel for the appellant made the undertaking (now fulfilled) that a dividend be paid to non-corporate shareholders so as to allow the matter to be resolved at this juncture. With the agreement of the Crown, the Court agreed to conduct the abuse analysis on the basis that a tax benefit arose by reason of such a dividend having been paid.
[50] The question that arises in that context is whether the series of transactions which allowed for the payment of a capital dividend for the benefit of the non-corporate shareholders, equal to the whole of the capital gain realized on the sale of the real estate assets, frustrates the provision of the Act that were relied upon in order to achieve this result.
B. Standard of Review
[51] The question whether there has been an abuse is one of mixed fact and law (Canada Trustco, para. 44; Housen v. Nikolaisen, 2002 SCC 33, [2002] 2 S.C.R. 235 [Housen], para. 37). The abuse analysis requires in turn that the object, spirit and purpose of the provisions enabling the tax benefit be determined (Copthorne, para. 70), an exercise that gives rise to an extricable question of law to be assessed on a standard of correctness (Canada Trustco, para. 44; Housen, paras. 8, 37).
C. Statutory Interpretation in the Context of the GAAR
[52] The Tax Court judge properly instructed himself as to the exercise that must be undertaken in construing the relevant provisions in a GAAR context (Reasons, para. 33):
The first step involves identifying the object, spirit and purpose of the relevant rules. Statutory interpretation under the GAAR differs from traditional word-based interpretation. Whereas, under the traditional rule of statutory interpretation, the analysis seeks to determine what the meaning of a provision is, under the GAAR, statutory interpretation is used to determine the object, spirit or purpose of the provision. The object, spirit or purpose is the rationale underlying the provision. Transactions may be found to be abusive of a provision’s underlying rationale, even though they are consistent with the literal, contextual and purposive meaning of the words of the statute. [Copthorne, para. 70 and Canada v. Oxford Properties Group Inc., 2018 FCA 30, [2018] 4 F.C.R. 3 [Oxford], paras. 40-46 are cited in support.]
[53] It follows that a GAAR analysis can lead to a result that is different from that obtained by a traditional, textual, contextual and purposive interpretation focused on the meaning of the words. Indeed, this must be so given that (Copthorne, para. 109):
[w]hen the Minister invokes the GAAR, he is conceding that the words of the statute do not cover the series of transactions at issue. Rather, he argues that although he cannot rely on the text of the statute, he may rely on the underlying rationale or object, spirit and purpose of the legislation to support his position.
And that:
if the Court [was] confined to a consideration of the language of the provisions in question, without regard to their underlying rationale, it would seem inevitable that the GAAR would be rendered meaningless [Copthorne, para. 111, citations omitted].
D. Statutory Context
[54] Because we are concerned here with provisions that deal with partnership interests—specifically the manner in which their ACB is computed—it is useful to have in mind in broad terms the tax treatment of partnerships under the Act. As was explained in Oxford (para. 48):
[p]artnerships have a hybrid status under the Act. Although partnership income is allocated to the partners, it is computed “as if the partnership were a separate person” (paragraph 96(1)(a)). Because partnerships are distinct from the partners at the income computation stage – Division B – computation of income – they, much like corporations, can hold assets, in which case the interest of the partners in those assets is reflected by their partnership interests. Partnership interests are distinct from the underlying property held by the partnership and can be subject to a different treatment under the Act.
[55] While the text of subsection 245(4) provides that abusive tax avoidance is to be determined “having regard to [the] provisions [of the Act], […] read as a whole,” it is appropriate to circumscribe the analysis to provisions that are “grouped together” or “work together to give effect to a plausible and coherent plan” (Copthorne¸ para. 91 citing R. Sullivan, Sullivan on the Construction of Statutes (5th ed. 2008), at pp. 361 and 364). Two distinct groups of related provisions interact in the present case, the first relating to the CDA regime and the second to the negative ACB rules.
E. The Object, Spirit and Purpose of the Relevant Provisions
I. Subsections 83(2), 89(1) and 184(2): the CDA Regime
[56] A CDA is a notional account maintained by private corporations to keep track of certain types of tax-free surpluses accumulated over time (Reasons, para. 39). As per the definition found at subsection 89(1), the balance of the CDA is determined at any particular time by adding, inter alia, (i) the tax-free portion of capital gains, (ii) the amount of tax-free capital dividends received by the corporation from other corporations and (iii) the proceeds of certain life insurance policies and subtracting, inter alia, (iv) the non-deductible portion of capital losses and (v) capital dividend distributions made before the particular time (Reasons, para. 39).
[57] Because a CDA is computed by reference to the qualifying amounts that may be distributed tax-free, it must, by definition, reflect a positive amount. Mathematically, however, amounts that impact negatively on the CDA are kept track of even when they bring the balance below zero in which case the CDA cannot become positive again unless and until the negative balance is compensated by additional qualifying amounts specified in subsection 89(1). Of significance for present purposes is that should a negative balance be caused by a capital loss that offsets a capital gain that has already been used to make a tax-free distribution, the system is kept whole and its integrity is preserved by the fact that the CDA deficit will have to be compensated by additional qualifying amounts before a capital dividend can again be paid.
[58] When a private corporation has a positive CDA balance, it may distribute those surpluses, tax-free, by way of a capital dividend, but only to the extent of the corporation’s CDA balance immediately before the dividend becomes payable (subsection 83(2)). Should a corporation elect to pay a capital dividend in excess of the CDA, it incurs the additional tax on excessive elections imposed under Part III of the Act, specifically subsection 184(2), unless it elects, with the agreement of the relevant shareholders, to treat the excess as a regular taxable dividend (subsection 184(3) and 184(4)).
[59] As explained by the Tax Court judge, the CDA regime plays a fundamental role in giving effect to the principle of integration by ensuring that income is “taxed at the same rate whether it is earned directly or indirectly by a corporation” (Reasons, para. 41).
[60] In broad terms, the CDA regime seeks to achieve this result by neutralizing the impact of the interposition of a corporation in the manner in which capital gains are taxed. Given that only one half of capital gains is taxable (section 38), Parliament provided for a mechanism whereby a corporation can preserve the tax-free portion of the gain for distribution to a shareholder without attracting an extra level of tax—this mechanism governs the manner in which the CDA is computed. In essence, the CDA regime ensures that no more than the tax-free portion is distributed to shareholders by way of a capital dividend so as to mirror the tax treatment of an individual taxpayer who generates the underlying gain directly.
[61] The same rationale governs the tax treatment of capital losses. For that purpose, when a corporation suffers a capital loss, a portion of the loss that corresponds with the non-taxable capital gain portion is deducted from the CDA, thereby lowering the amount available for capital dividend election and distribution. This again mimics the effect of a capital loss when incurred by an individual taxpayer directly. The CDA computation mechanism reflects this by decreasing the amount that can be paid out tax-free by a corresponding amount whenever a private corporation suffers a capital loss.
[62] As pointed out by the Tax Court judge, although the CDA regime does not achieve integration in all instances or perfectly, this does not detract from the fact that subsections 83(2), 89(1) and 184(2) were enacted to promote integration. Private corporations must keep track of the tax-free portion of capital gains and the non-deductible portions of capital losses, determine how and when capital dividends may be paid out, and suffer penalties when excessive dividends are paid out (Reasons, para. 46).
II. Subsections 40(3.1) and 40(3.12): the Negative ACB Rules
[63] The 1994 Budget Supplementary Information presents subsection 40(3.1) as an anti-avoidance provision which was enacted in the following context (p. 42):
The adjusted cost base (ACB) of a taxpayer’s property reflects the cost to a taxpayer of the property and is taken into account in computing a capital gain or loss on the sale of the property. In certain circumstances the ACB of a taxpayer’s property may become negative in which case the taxpayer is treated as having realized a capital gain. This rule generally does not apply where the property is a partnership interest; this exception recognizes that a partner’s negative ACB may result from legitimate, and possibly temporary, circumstances such as the withdrawal of partnership capital or where losses of the partnership are allocated to the partner for tax purposes.
Certain tax shelters have been structured to utilize this exception to permit limited or passive investors to claim tax-deductible losses and receive cash distributions which exceed the amount invested: that is, to extract tax-free from the partnership more than the cost of the partnership interest to the investor. The budget proposes that limited and certain other passive partners will be required to report as a capital gain any negative ACB in their partnership interest at the end of a fiscal period of the partnership.
These rules are a logical extension of the existing limited partnership “at-risk” rules which constrain the amount of losses that may be flowed out to an investor. In particular, the budget proposal will ensure that the at-risk rules cannot be circumvented by having partnership loss allocations precede distributions. Subject to transitional provisions, the new rules will apply to fiscal periods of partnerships ending after February 22, 1994.
[64] Prior to the 1994 amendments, partnership interests were excluded from the application of subsection 40(3), which deems a capital gain to arise when the ACB of capital property becomes negative. The application of this rule to partnership interests was considered to be ineffective and cumbersome given that partnerships often make distributions from available cash as opposed to accumulated net income (Reasons, para. 50). However, the exclusion of partnership interests from the application of subsection 40(3) became a source of concern with the advent of tax sheltered partnership investments, a development which led to the 1994 amendments (Reasons, paras. 52-56). The negative ACB rules were said to be a logical extension of the partnership “at-risk” rules (Reasons, para. 57 citing 1994 Budget Supplementary Information at p. 42). These amendments bring partnership interests within the scope of the negative ACB rules (subsection 40(3.1)) but also provide relief by allowing the deemed gain that arises when the ACB of partnership interests become negative to be offset by a deemed loss up to an elected amount not to exceed the deemed gain (subsection 40(3.12)). When regard is had to the circumstances which led to the adoption of these provisions, subsection 40(3.1) is a specific anti-avoidance measure which levies tax and subsection 40(3.12) provides corresponding relief.
[65] The mechanical operation of these provisions is premised on paragraph 53(2)(c) which provides that the ACB of a partner’s partnership interest is to be reduced by an amount equal to that of the distribution made to it by the partnership. Where, at the end of the fiscal period, the partner is a limited partner or a specified member of the partnership (mostly passive partners) and the amount of this distribution exceeds the ACB leading it to be negative, the Act provides for a deemed capital gain in the amount of the negative ACB so as to bring it back to 0 (subsections 40(3.1) and 40(3.11)). In a subsequent fiscal period, when the ACB of the partnership interest becomes positive again, the partner can elect to have a deemed capital loss of an amount equal to the deemed capital gain triggered in a previous fiscal period, but not exceeding the current ACB (subsection 40(3.12)). The deferral of this relief to a subsequent fiscal period results from the fact that the negative ACB triggering the deemed gain is to be computed at the close of a fiscal period.
[66] From a tax policy perspective, the impact of subsections 40(3.1) and 40(3.12) was to be a wash as it was expected that, the ACB of their partnership interest permitting, partners would elect to recover the tax that they had to pay on the deemed capital gain arising in a prior year. For the same reason, the corresponding upward and downward impact that these provisions have on the CDA was expected to self-erase.
F. The Abuse Analysis
[67] Two preliminary comments are in order. In resisting the appeal, the Crown asserts, with reference to subsection 40(3.1), that “[t]he use of an anti-avoidance provision to obtain a tax benefit […] is abusive tax planning under the GAAR” (Memorandum of the Crown, Overview, p. 1). Lipson and Satoma Trust are cited in support of this proposition.
[68] I do not believe that these decisions support so broad a proposition. As Lipson makes clear, the abuse lies in using an anti-avoidance measure in order to obtain the result that it is intended to circumvent. In that case, subsection 74.1(1), a provision that was intended to prevent spouses from reducing their tax burden by taking advantage of their non-arm’s length relationship, was shown to have been used to achieve a tax reduction (Lipson, para. 42). Similarly, in Satoma Trust, an anti-avoidance measure that was intended to prevent income splitting (subsection 75(2)) was used in combination with another provision (subsection 112(1)) to achieve income splitting (Satoma Trust, para. 52).
[69] As is the case for any other provision, what must be shown at the abuse stage of the analysis is that the anti-avoidance provision was used in a manner that defeats its underlying rationale. In the present case, I agree with the appellant that subsection 40(3.1) operated exactly as it should when regard is had to its underlying rationale (see para. 34 above).
[70] I also agree with the appellant, based on the binding decisions that it cites (see para. 35 above), that in conducting the abuse analysis, the Court must give effect to the transactions as they unfolded, and refrain from assessing the abuse on th