Citation: 2013 TCC 86
HER MAJESTY THE QUEEN,
REASONS FOR JUDGMENT
 The Minister of National Revenue (the "Minister") reassessed Brianne and Steven Gwartz’s 2003, 2004 and 2005 taxation years. In the reassessments, the Minister relied on the general anti-avoidance rule (the "GAAR") contained in section 245 of the Income Tax Act (Canada) (the "ITA") to recharacterize as dividends certain capital gains which had been realized by a family trust and allocated to the appellants in 2003, 2004 and 2005. This in turn gave rise to the imposition of the tax on "split income", provided for in section 120.4 of the ITA, on the income recharacterized as dividend income. The appeals were heard on common evidence.
 The capital gains allocated to the appellants were realized by a family trust after the trust sold certain shares of a management corporation that acted for the dental practice of the appellants’ father. These shares, which had a low paid-up capital and a high redemption value, had been received by the family trust by way of a stock dividend paid by the management corporation.
 The appellants concede the existence of a "tax benefit" and an "avoidance transaction", hence the existence of "abusive tax avoidance" is the only issue before the Court. The respondent argues that the appellants abusively circumvented section 120.4 such that the GAAR should apply.
II. FACTUAL BACKGROUND
 The parties filed an agreed statement of facts which is summarized below.
 Dr. Mark Gwartz, the appellants’ father, is a dentist. Forest Hill Dental Management Inc. ("FHDM") is a management corporation that acted for Dr. Gwartz’s dental practice during the relevant period. Certain shares of FHDM were held by the Gwartz/Ludwig Family Trust (the "Trust"), which was settled in 1994 by Dr. Gwartz’s mother, Bernice Gwartz. The appellants were both beneficiaries of the Trust at all relevant times, as were certain other family members.
 Prior to the transactions in issue, the Trust held all of the common shares and all of the Class C Preferred Shares in FHDM. On December 31, 2003, FHDM issued 150,000 preferred shares of a new class, Class D, to the Trust as a stock dividend on the common shares held by the Trust (the Class D Preferred Shares). The Class D Preferred Shares were each redeemable and retractable for $1, but, pursuant to the corporate resolution that authorized the payment of the stock dividend, only $1 in total was added to the stated capital account maintained for that class. In other words, the Class D Preferred Shares had a high redemption value and a low paid-up capital – they were so-called "high-low" shares. The effect of the above transaction was to shift value from the common shares to the Class D Preferred Shares. As a result, part of the gain accrued up until that time on the common shares held by the Trust was transferred to the Class D Preferred Shares.
 On the same day, the Trust sold 75,000 of the Class D Preferred Shares to Dr. Gwartz in exchange for an interest-bearing promissory note in the amount of $75,000. The Trust sold its remaining 75,000 Class D Preferred Shares to Dr. Gwartz on December 15, 2004, also in exchange for an interest-bearing promissory note in the amount of $75,000.
 On January 15, 2005, FHDM issued a further 150,000 Class D Preferred Shares to the Trust, which were each redeemable and retractable for $1, and added $1 in total to the stated capital account maintained for that class. On January 30, 2005, the Trust sold 75,000 of the Class D Preferred Shares to Dr. Gwartz in exchange for an interest-bearing promissory note in the amount of $75,000.
 Dr. Gwartz subsequently sold his 225,000 Class D Preferred Shares in FHDM to 2062067 Ontario Inc. (“2062067”), a corporation wholly owned by his spouse, in return for an interest-bearing promissory note in the amount of $225,000.
 On February 1, 2005, FHDM redeemed the Class D Preferred Shares held by 2062067 for $225,000. In its tax return, 2062067 reported a deemed dividend of $224,999 in respect of that redemption, and claimed an offsetting deduction under section 112 of the ITA. The proceeds from the redemption were used by 2062067 to extinguish its promissory note in favour of Dr. Gwartz, and Dr. Gwartz then used the $225,000 to extinguish his promissory notes in favour of the Trust (whose principal amounts totalled $225,000).
 For each of the 2003, 2004 and 2005 taxation years, the Trust reported a $74,999.50 capital gain on the sale of 75,000 Class D Preferred Shares in FHDM, and allocated those gains entirely to the appellants. In the 2003 and 2004 taxation years, those capital gains were allocated equally between the appellants such that each was allocated $37,499.75 in each year. In the 2005 taxation year, the Trust allocated $24,999.83 to Steven and $49,999.67 to Brianne. The appellants reported the capital gains allocated to them in their tax returns.
 Steven Gwartz turned 17 during 2004 and Brianne Gwartz was under the age of 17 at all material times.
 The Minister, relying on the GAAR, reassessed the appellants in respect of their 2003, 2004 and 2005 taxation years. According to the letter explaining the reassessments, the Minister relied on the GAAR to recharacterize the capital gains as dividend income. This resulted in the tax on “split income” provided for in section 120.4 of the ITA (colloquially known as the “kiddie tax”) applying to the recharacterized income.
(1) Evidentiary Issue: Admissibility of the CRA Documents
 A preliminary issue relates to two internal Canada Revenue Agency (the "CRA") documents that the appellants seek to introduce as evidence. These documents are a memorandum from the CRA’s Hamilton Tax Services Office to the CRA’s GAAR & Technical Support Section concerning the applicability of the GAAR (the “GAAR Committee Referral”), and a T401 Report on Objection which was prepared by the CRA in respect of Brianne Gwartz’s Notice of Objection (the “T401”, together with the GAAR Committee Referral, the “CRA Documents”). The respondent disclosed the CRA Documents during the discovery process.
 Counsel for the appellants argues that the CRA Documents are relevant because they illustrate why the Minister thought it was appropriate to apply the GAAR to the transactions in issue.
(2) Substantive Issue: Application of the GAAR
 The appellants concede that: (i) the transactions in issue constituted a “series of transactions” within the meaning of subsection 248(10) of the ITA; (ii) this series of transactions gave rise to “tax benefits” as defined in subsection 245(1) of the ITA; and (iii) one or more of the transactions forming part of the series of transactions constituted “avoidance transactions” as that term is defined in subsection 245(3) of the ITA. As a result, the appellants’ submissions were restricted to the issue of whether there was a misuse or abuse of the ITA for the purposes of section 245(4) of the ITA.
 The appellants rely on the framework established in Canada Trustco Mortgage Co. v. Canada and Copthorne Holdings Ltd. v. Canada with respect to the possible application of subsection 245(4). The appellants submit that the object, spirit and purpose of section 120.4 were not frustrated by the transactions in issue. Counsel for the appellants admits that the transactions in issue did reflect deliberate tax planning, but argues that deliberate tax planning is not the same as abusive tax avoidance.
 With respect to the text of section 120.4, counsel suggests that the provision applies only to certain types of persons and income. With respect to the context of section 120.4, counsel submits that other provisions in the ITA specifically address the extraction of corporate surplus other than by way of a dividend. For example, counsel cites subsection 15(1.1), which, in certain circumstances, deems the fair market value of the stock dividends, in transactions involving the payment of stock dividends, to be included in income. Such transactions might otherwise give rise to amounts that would be taxed on capital account. Counsel for the appellant argues that, in light of such other specific rules contained in the ITA, the fact that section 120.4 did not contain a specific rule deeming certain capital transactions, such as the ones in issue, to give rise to dividends illustrates a deliberate choice by Parliament not to extend section 120.4 to capital gains transactions which have as one of their purposes or results the extraction of corporate surpluses.
 With respect to the purpose of section 120.4, the appellants rely on certain extrinsic materials and similarly conclude that Parliament made the conscious choice to enact a narrowly targeted rule that excluded capital gains.
 The appellants also submit that recent case law supports their position. They cite McClarty Family Trust v. The Queen, in which Angers J. held that the GAAR did not apply to transactions that were (in the words of the appellants) “virtually identical” to those in issue in these appeals. In particular, they note Angers J.’s comments in obiter that it is inappropriate for the Minister to use the GAAR to fill gaps that Parliament may have left when enacting section 120.4. The appellants cite a passage from Lehigh Cement Limited v. Canada, in which Sharlow J.A. held that “the Crown cannot discharge the burden of establishing that a transaction results in the misuse of an exemption merely by asserting that the transaction was not foreseen or that it exploits a previously unnoticed legislative gap”.
 The appellants argue that the ITA does not contain an overarching policy against surplus stripping. Counsel for the appellants further contends that there is in the ITA no overarching policy against income splitting either. Counsel observes that the ITA even promotes income splitting in certain circumstances.
 In addition, the appellants submit that the transactions at issue were not artificial and did not give rise to mere “paper losses” (or, rather, “paper gains”), as in Triad Gestco Ltd. v. The Queen. Instead, the appellants contend that the transactions were carried out in reliance on a well-known tax planning technique to ensure that they were principally taxed on real taxable capital gains rather than taxable dividends.
 Finally, the appellants argue that the transactions in issue did not misuse or abuse any provisions of the ITA. The appellants submit that the tax results of the transactions in issue are the consequence of highly specific rules set out in the ITA, and that each of the provisions relied upon operated exactly as intended.
(1) Evidentiary Issue: Admissibility of the CRA Documents
 Counsel for the respondent admits that the CRA Documents are authentic. In addition, counsel consents to the T401 being admitted into evidence for informational purposes (in counsel’s words, “the Court can look at it and say, ‘Here is the T401’”). Indeed, counsel for the respondent referred to the contents of the T401 during her oral submissions. Counsel for the respondent also appears to accept that the GAAR Committee Referral would be admissible under section 100 of the Tax Court of Canada Rules (General Procedure) because it was disclosed as part of the discovery process. This implies that the respondent accepts that the GAAR Committee Referral “is otherwise admissible”, which is a condition that must be met before section 100 can apply.
 The respondent’s position appears to be that both of the CRA Documents can be admitted as evidence as to the Minister’s analysis in confirming the reassessments and deciding to apply the GAAR, but not as evidence for the truth of their contents. Nevertheless, counsel for the respondent also argues that the CRA Documents are hearsay, irrelevant, have no probative value and should be given no weight.
(2) Substantive Issue: Application of the GAAR
 The respondent concedes that the reassessment in respect of Steven’s 2005 taxation year should be vacated. Steven turned 17 during 2004 so that he was not a “specified individual”, as that term is defined in subsection 120.4(1), during his 2005 taxation year. As a result, the respondent acknowledges that the capital gain allocated to him in that year by the Trust cannot be recharacterized as a dividend.
 The respondent’s position is that the Minister properly applied the GAAR because the transactions in issue (other than those transactions relating to Steven’s 2005 taxation year) circumvented the application of section 120.4 of the ITA in a manner that frustrated or defeated the object, spirit or purpose of that provision. The respondent submits that the purpose of section 120.4 is to prevent income splitting with minors. Counsel for the respondent suggests, more specifically, that section 120.4 is targeted at income splitting with minorswhere those minors receive certain types of income that are susceptible to manipulation. As support for this proposition, counsel relies on a publication by the Department of Finance released in conjunction with the 1999 Budget, which states: “Dividends received on any listed shares will not be subject to these rules, since the income flow is less susceptible to manipulation”. In addition, counsel submits that the policy underlying section 120.4 has always reflected Parliament’s concern with income splitting involving minors where shares of private companies are involved.
 The respondent argues that capital gains were not included in the initial version of section 120.4 because, at the time, capital gains were not identified as the type of income that section 120.4 was designed to tax. Counsel suggests that Parliament designed the provision to apply to then current income splitting techniques, but intended to monitor the effectiveness of the measure and take appropriate action if new techniques were developed. The respondent contends that Parliament did not foresee the use of “artificial” capital gains to engage in income splitting with minors. The use of capital gains represents an evolution of the tax‑planning techniques used in that context.
 Counsel for the respondent also argues that that the 2011 amendments to section 120.4, which, as discussed below, apply to certain capital gains, did not reflect a change in policy. Rather, counsel argues that the amendments reflected a move by Parliament to close a “loophole”. Counsel compared the situation to that which was considered by the Federal Court of Appeal in Water's Edge Village Estates (Phase II) Ltd. v. Canada. In that decision, Noël J.A. commented that a subsequent amendment demonstrated that “Parliament moved as quickly as it could to close the loophole”.
 Counsel also noted that, in McClarty Family Trust, Angers J. recently commented on the applicability of the GAAR to transactions that are similar to those before the Court in these appeals. After finding that the GAAR did not apply because of the absence of an avoidance transaction, Angers J. stated that “[t]here is a definite gap that was left by Parliament in enacting section 120.4 of the ITA”, and cited Landrus v. The Queen in support of the proposition that it is inappropriate to use the GAAR to fill such gaps. Counsel for the respondent argues that this aspect of the McClarty Family Trust decision should not be followed, in part because it was obiter. In addition, counsel cites the Copthorne decision (which, counsel notes, was released after the hearing in the McClarty Family Trust case), in which, in relation to the relevance of the principle of “implied exclusion” in the GAAR context, Rothstein J. wrote:
. . . When 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.
 Finally, the respondent argues that the transactions in issue artificially created capital gains. At the hearing, counsel argued that the transactions in issue had the same circularity and artificiality as the transactions at issue in Triad Gestco and 1207192 Ontario Limited v. The Queen. Counsel suggests that the transactions do not reflect an increase in real economic power, but merely effected a shift of value within the family unit. Counsel also suggests that the capital gains received by appellants were manufactured in order to convert dividends into capital gains.
 The respondent’s submissions, as reflected in both the Written Submissions of the respondent and the oral submissions of counsel at trial, were directed at the question of abusive circumvention of section 120.4. However, in the respondent’s reply to the notice of appeal in respect of each appellant it is submitted at paragraph 19 that:
The Transactions may reasonably be considered to have resulted directly or indirectly in a misuse of a provision of the Act, including but not limited to subsection 84(3) and sections 38, 39, 40, 82, 84 and 120.4 of the Act or an abuse having regard to the provisions of the Act read as a whole, within the meaning of subsection 245(4) of the Act.
In addition, in the reply, the respondent submitted that the ITA, read as a whole, is designed to prevent the extraction of corporate surplus on a tax-free or tax-reduced basis.
 At the hearing, counsel clarified that the respondent is no longer pleading misuse or abuse of subsection 84(3) of the ITA or the existence of an overall scheme in the ITA against surplus stripping.
IV. ISSUE TO BE DECIDED
 The issue to be decided is whether section 245 of the ITA properly applies so as to allow the recharacterization as dividend income of the capital gains that the Trust allocated to Brianne in her 2003, 2004 and 2005 taxation years, and to Steven in his 2003 and 2004 taxation years (the “Relevant Capital Gains”). As discussed above, the appellants concede the existence of “tax benefits” within the meaning of subsection 245(1) and “avoidance transactions” within the meaning of subsection 245(3). The applicability of the GAAR therefore turns on whether there was “misuse” or “abuse” for the purposes of subsection 245(4).
 At the hearing, I reserved my decision on the admissibility of, and weight that should be given to, the CRA Documents. As noted, these documents are a memorandum to the CRA’s GAAR & Technical Support Section in respect of the applicability of the GAAR, and a T401 report on objection which was prepared by the CRA in respect of Brianne Gwartz’s notice of objection. Each of the CRA Documents essentially contains a description of the facts in issue together with an analysis of the GAAR and its applicability. The T401 additionally discusses how the transactions in issue should be treated under the CRA’s administrative positions.
 As noted above, counsel for the respondent apparently consents to the admission of each of the CRA Documents for informational purposes, but argues that they are hearsay, irrelevant, have no probative value and should be given no weight.
 Evidence that lacks relevance to the matters in dispute, as framed in the pleadings, is inadmissible. In this case, the parties have agreed on all of the material facts. Moreover, the assumptions made by the Minister when reassessing the appellants are not in issue. Therefore, the CRA Documents lack relevance.
(1) The Three-Step Framework from Canada Trustco
 In Canada Trustco, the Supreme Court established a three-step framework for determining whether the GAAR applies to a transaction or series of transactions. This framework was reasserted by the Supreme Court in Mathew v. Canada, Lipson v. Canada, and Copthorne.
 Within this framework, the first step is to inquire into the existence of a “tax benefit” within the meaning of subsection 245(1). For there to be a tax benefit, a transaction, or series of transactions of which the transaction is a part, must result in “a reduction, avoidance or deferral of tax or other amount . . . payable” under the ITA or other relevant source of tax law, or “an increase in a refund of tax or other amount” under the ITA or other relevant source of tax law. In this case, the appellants concede the existence of a tax benefit. While there was some debate at the hearing regarding the nature of the tax benefit conceded by the appellants – that is whether it resulted from the avoidance of the tax imposed by section 120.4, from the comparatively low marginal income tax rates paid by the appellants, or from the fact that the appellants received capital gains rather than income – the answer is ultimately unimportant because the issue was conceded.
 Under the second step of the framework established in Canada Trustco, the transaction giving rise to the tax benefit must be an “avoidance transaction” within the meaning of subsection 245(3). In this case, the appellants concede the existence of an avoidance transaction.
 These appeals thus turn on the outcome of the third step of the framework established in Canada Trustco, which involves a determination whether the avoidance transaction giving rise to the tax benefit is abusive under subsection 245(4). Within this framework, the abuse inquiry involves, first, interpreting the relevant provisions of the ITA to determine their object, spirit and purpose and, second, determining whether the impugned transactions fall within, or frustrate the purpose of those provisions. As described in Copthorne:
69 In order to determine whether a transaction is an abuse or misuse of the Act, a court must first determine the “object, spirit or purpose of the provisions. . . that are relied on for the tax benefit, having regard to the scheme of the Act, the relevant provisions and permissible extrinsic aids” (Trustco, at para. 55). The object, spirit or purpose of the provisions has been referred to as the “legislative rationale that underlies specific or interrelated provisions of the Act” (V. Krishna, The Fundamentals of Income Tax Law (2009), at p. 818).
70 The object, spirit or purpose can be identified by applying the same interpretive approach employed by this Court in all questions of statutory interpretation — a “unified textual, contextual and purposive approach” (Trustco, at para. 47; Lipson v. Canada, 2009 SCC 1,  1 S.C.R. 3, at para. 26). While the approach is the same as in all statutory interpretation, the analysis seeks to determine a different aspect of the statute than in other cases. In a traditional statutory interpretation approach the court applies the textual, contextual and purposive analysis to determine what the words of the statute mean. In a GAAR analysis the textual, contextual and purposive analysis is employed to determine the object, spirit or purpose of a provision. Here the meaning of the words of the statute may be clear enough. The search is for the rationale that underlies the words that may not be captured by the bare meaning of the words themselves. However, determining the rationale of the relevant provisions of the Act should not be conflated with a value judgment of what is right or wrong nor with theories about what tax law ought to be or ought to do.
 An arrangement will be abusive if it circumvents the application of a specific anti-avoidance rule or provision that is relied upon in a manner that frustrates the object, spirit and purpose of the rule.
 The existence of abusive tax avoidance must be clear. If it is not, the benefit of the doubt goes to the taxpayer. Moreover, the Minister bears the burden of establishing abusive tax avoidance.
 Before undertaking the two-stage analysis for abusive tax avoidance mandated by Canada Trustco, I will review certain principles in relation to (i) tax planning in general, (ii) the appropriateness of using the GAAR as a gap-filling measure, (iii) the existence of a general policy in the ITA regarding surplus stripping, (iv) the existence of a general policy in the ITA regarding income splitting, and (v) the significance in a GAAR analysis of subsequent amendments to a provision purportedly abused. All of these elements have a direct bearing on the GAAR analysis in these appeals.
(2) Tax Planning Is Not Inherently Abusive
 Canada Trustco and Copthorne each reiterate the principle that tax planning is not per se abusive for the purposes of subsection 245(4). In Canada Trustco, the Court stated:
61 A proper approach to the wording of the provisions of the Income Tax Act together with the relevant factual context of a given case achieve balance between the need to address abusive tax avoidance while preserving certainty, predictability and fairness in tax law so that taxpayers may manage their affairs accordingly. Parliament intends taxpayers to take full advantage of the provisions of the Act that confer tax benefits. Parliament did not intend the GAAR to undermine this basic tenet of tax law.
Similarly, in Copthorne, the Court said that “[t]axpayers are entitled to select courses of action or enter into transactions that will minimize their tax liability”.
 As a result, a taxpayer who chooses a course of action that minimizes his or her tax liability will not necessarily have engaged in abusive tax avoidance for the purposes of subsection 245(4).
(3) Gap Filling and the GAAR
 Abusive tax avoidance cannot be found to exist if a taxpayer can only be said to have abused some broad policy that is not itself grounded in the provisions of the ITA. In Canada Trustco, the Court stated:
41 The courts cannot search for an overriding policy of the Act that is not based on a unified, textual, contextual and purposive interpretation of the specific provisions in issue. First, such a search is incompatible with the roles of reviewing judges. The Income Tax Act is a compendium of highly detailed and often complex provisions. To send the courts on the search for some overarching policy and then to use such a policy to override the wording of the provisions of the Income Tax Act would inappropriately place the formulation of taxation policy in the hands of the judiciary, requiring judges to perform a task to which they are unaccustomed and for which they are not equipped. Did Parliament intend judges to formulate taxation policies that are not grounded in the provisions of the Act and to apply them to override the specific provisions of the Act? Notwithstanding the interpretative challenges that the GAAR presents, we cannot find a basis for concluding that such a marked departure from judicial and interpretative norms was Parliament’s intent.
42 Second, to search for an overriding policy of the Income Tax Act that is not anchored in a textual, contextual and purposive interpretation of the specific provisions that are relied upon for the tax benefit would run counter to the overall policy of Parliament that tax law be certain, predictable and fair, so that taxpayers can intelligently order their affairs. Although Parliament’s general purpose in enacting the GAAR was to preserve legitimate tax minimization schemes while prohibiting abusive tax avoidance, Parliament must also be taken to seek consistency, predictability and fairness in tax law. These three latter purposes would be frustrated if the Minister and/or the courts overrode the provisions of the Income Tax Act without any basis in a textual, contextual and purposive interpretation of those provisions.
A related principle is that it is inappropriate, where the transactions do not otherwise conflict with the object, spirit and purpose of the provisions of the ITA to apply the GAAR to deny a tax benefit resulting from a taxpayer’s reliance on a previously unnoticed legislative gap. This principle is illustrated in the decision of the Federal Court of Appeal in Lehigh Cement. The Tax Court of Canada had dismissed the taxpayer’s appeal in respect of the application of the GAAR. Sharlow J.A., for the Federal Court of Appeal, in ruling that the GAAR did not apply, as follows:
37 When Parliament adds an exemption to the Income Tax Act, even one as detailed and specific as subparagraph 212(1)(b)(vii), it cannot possibly describe every transaction within or without the intended scope of the exemption. Therefore, it is conceivable that a transaction may misuse a statutory exemption comprised of one or more bright line tests such as, in this case, the arm’s length test and the 5 year test. However, the fact that an exemption may be claimed in an unforeseen or novel manner, as may have occurred in this case, does not necessarily mean that the claim is a misuse of the exemption. It follows that the Crown cannot discharge the burden of establishing that a transaction results in the misuse of an exemption merely by asserting that the transaction was not foreseen or that it exploits a previously unnoticed legislative gap. As I read Canada Trustco, the Crown must establish by evidence and reasoned argument that the result of the impugned transaction is inconsistent with the purpose of the exemption, determined on the basis of a textual, contextual and purposive interpretation of the exemption.
 My colleague Paris J. adopted a similar view in Landrus:
124 The Minister is therefore using the GAAR in this case to fill in the gaps left by Parliament in subsection 85(5.1). This is an inappropriate use of the GAAR, as noted by Bowman A.C.J. in Geransky v. R.:
...The Income Tax Act is a statute that is remarkable for its specificity and replete with anti-avoidance provisions designed to counteract specific perceived abuses. Where a taxpayer applies those provisions and manages to avoid the pitfalls the Minister cannot say “Because you have avoided the shoals and traps of the Act and have not carried out your commercial transaction in a manner that maximizes your tax, I will use GAAR to fill in any gaps not covered by the multitude of specific anti-avoidance provisions”.
(4) Is There a Policy in the ITA Against Surplus Stripping?
 Surplus stripping involves the extraction of corporate surplus in a manner other than by way of a dividend, usually by way of a capital gain. For example, rather than arranging for a corporation to pay out its retained earnings as a dividend, a shareholder might sell shares in the capital stock of the corporation to a related corporation. Such a transaction would be advantageous if the resulting capital gain were subject to less tax than the dividend would have been. Surplus stripping is “[o]ne of the most longstanding and persistent sources of conflict between taxpayers and tax collectors.”
 The courts have held that surplus stripping does not inherently constitute abusive tax avoidance. In Collins & Aikman Products Co. et al. v. The Queen, Boyle J. wrote that:
77 Similarly, Campbell J. in Copthorne Holdings Ltd. v. The Queen, 2007 DTC 1230, wrote, at paragraph 73:
While the Act contains many provisions which seek to prevent surplus stripping, the analysis under subsection 245(4) must be firmly rooted in a unified textual, contextual and purposive interpretation of the relevant provisions. As such, reliance on a general policy against surplus stripping is inappropriate to establish abusive tax avoidance.
78 Also to like effect, Lamarre J. in McMullen v. The Queen, 2007 DTC 286, wrote at paragraph 56:
In conclusion, the respondent has not persuaded me, or has not presented any evidence establishing, that there was any abuse of the Act read as a whole, or that the policy of the Act read as a whole is designed so as to necessarily tax corporate distributions as dividends in the hands of shareholders. In any event, as the Supreme Court of Canada has said, “[i]f the existence of abusive tax avoidance is unclear, the benefit of the doubt goes to the taxpayer”. . .
79 The words of Bowman C.J., Campbell J. and Lamarre J. apply equally in this case.
Similarly, Rothstein J. held in Copthorne that, in determining whether there has been abusive tax avoidance, “[w]hat is not permissible is basing a finding of abuse on some broad statement of policy, such as anti-surplus stripping, which is not attached to the provisions at issue.”
 It is also noteworthy that the respondent declined to pursue the position, put forward in each reply to the notice of appeal, that the appellants contravened a policy in the ITA against surplus stripping.
(5) Is There a Policy in the ITA Against Income Splitting?
 The ITA levies an income tax on individuals at marginal tax rates that increase as taxable income increases. Unlike the US personal income tax system, where married couples can file joint tax returns, the Canadian income tax system requires that couples and their children file individual tax returns in which each family member’s tax liability generally depends on that member’s individual circumstances. The increasing marginal tax rates and the choice of the individual as the basic taxable unit create incentives for taxpayers to split their income with their family members.
 With respect to the existence of a policy against income splitting, the Supreme Court of Canada in Neuman v. M.N.R. stated:
. . . [Subsection] 56(2) strives to prevent tax avoidance through income splitting; however, it is a specific tax avoidance provision and not a general provision against income splitting. In fact, “there is no general scheme to prevent income splitting” in the ITA (V. Krishna and J. A. Van Duzer, “Corporate Share Capital Structures and Income Splitting: McClurg v. Canada” (1992-93), 21 Can. Bus. L.J. 335, at p. 367). Section 56(2) can only operate to prevent income splitting where the four preconditions to its application are specifically met.
Second, this case concerns income received by Ruby Neuman during the 1982 taxation year at which time the ITA did not provide specific guidelines to deal with corporate structures designed for the purposes of income splitting and tax minimization. Professor V. Krishna, in an article entitled “Share Capital Structure of Closely-Held Private Corporations” (1996), 7 Can. Curr. Tax 7, at p. 9, made the following comment with respect to income splitting in the corporate context:
Except when specifically curtailed by the Income Tax Act (for example, by the attribution rules), income splitting per se is not a sanctioned arrangement. Thus, corporate structures that facilitate income splitting in private companies should not be penalized without clear statutory language and intent.
While Neuman predates the enactment of subsection 120.4, a broad policy in the ITA against income splitting, grounded in specific provisions of the ITA other than subsection 120.4, has not been recognized.
(6) The Significance of Subsequent Amendments
 The courts have taken differing approaches when considering the significance of subsequent amendments as an indicator of the policy underlying previous versions of a provision. In Water’s Edge, the Federal Court of Appeal dismissed in the following terms the appellant’s argument that the enacting of a subsequent amendment confirmed that the GAAR did not apply previously as follows:
46 Counsel for the appellants relied on the subsequent addition of subsection 96(8) to the Act. . . to argue that the transactions in issue do not offend any unwritten rule or policy. Subsection 96(8) was added by S.C. 1994, c. 21. . . and made applicable after December 21, 1992. Paragraph 96(8)(a) is of direct relevance. It specifically counters the result achieved by the appellants in this case by deeming the cost of acquisition of depreciable assets held by a foreign partnership to an incoming Canadian partner to be the lesser of its fair market value or its capital cost determined according to the ordinary rules.
47 Counsel argued that the prospective addition of subsection 96(8) demonstrates unequivocally that the transactions in issue did not offend the object and spirit of the Act at the time when they took place. I rather think that this amendment demonstrates that Parliament moved as quickly as it could to close the loophole exploited by the appellants precisely because the result achieved was anomalous having regard to the object and spirit of the relevant provisions of the Act.
Ultimately, the Federal Court upheld the Tax Court of Canada’s decision that the losses claimed by the appellants were properly denied by virtue of the GAAR.
 In contrast, one of the issues in each of Triad Gestco and 1207192 related to the Minister’s argument that the anti-avoidance rules in subparagraph 40(2)(g)(i) had been abusively circumvented. Subparagraph 40(2)(g)(i) is a “stop-loss” rule that can operate to deny a capital loss that is generated on a transaction between “affiliated persons”. The definition of “affiliated persons” in section 251.1 had been amended, effective after the impugned transactions took place. In Triad Gestco, Noël J.A. wrote:
56 I agree with Justice Paris that a reading of the relevant provisions does not support the existence of the policy identified by the Tax Court judge essentially for the reasons that he gave. When Parliament introduced the notion of “affiliated persons” back in 1995, it had to be aware that trusts could be used to counter the operation of subparagraph 40(2)(g)(i) and subsection 40(3.4). It is therefore reasonable to infer that a deliberate choice was made not to bring trusts within the definition. The fact that Parliament decided to alter this policy by including trusts on a prospective basis in 2005 cannot be relied on to infer that a policy to that effect was in place before the amendment (compare Water’s Edge Water’s Edge [sic]Village Estates (Phase II) Ltd. v. Canada, 2002 FCA 291,  2 F.C. 25, para. 47, where in contrast an amendment was held to be relevant because it had been enacted in order to close a blatant loophole).
 Paris J., in his reasons in 1207192, referred to by Noël J.A. in Triad Gestco, wrote:
76 The definition of “affiliated persons” in section 251.1 as it read for the year in issue sets out a carefully crafted group of relationships, and I believe that it is reasonable to infer that Parliament chose to limit the scope of the definition accordingly.
. . .
78 I held, though, that the amendments in Landrus did not “alter the fact that the stop-loss rules were exceptions that operate in well-defined circumstances” and that even after the amendments they did not deny losses on all transfers between all related parties. Therefore, the amendments were not material to the determination of the policy underlying the stop-loss rules in effect for the year under appeal.
 Those decisions demonstrate that a subsequent amendment, which would have defeated a tax avoidance strategy challenged under the GAAR, does not in itself indicate either that the strategy was abusive or that it was non-abusive. Instead, the subsequent amendment must be considered along with all other relevant materials to ascertain the object, spirit and purpose of the provision. In certain circumstances, a subsequent amendment might suggest that the provision’s object, spirit and purpose were frustrated by the tax avoidance strategy. In other circumstances, it might suggest that Parliament simply changed its mind and now intends to prevent something that initially was not intended to be captured by the provision.
(7) Do the Impugned Transactions Frustrate the Object Spirit and Purpose of Section 120.4?
(a) Text of Section 120.4
 Notwithstanding the fact that the wording of section 120.4 does not apply to the transactions in issue, that wording can shed light on the intent underlying the provision.
 Compared with other provisions in the ITA, the text of section 120.4 is notable for its relative brevity and simplicity. Its applicability or non-applicability to a transaction is, in general, readily observable. For example, determining whether a person is a specified individual, or determining whether certain dividend income received directly by a specified individual is split income, is straightforward. These factors suggest that, when enacting section 120.4, Parliament was concerned with minimizing the complexity of the provision and providing certainty to taxpayers with respect to its application.
(b) Context of Section 120.4
 In the ITA, certain provisions serve to eliminate the tax advantages that can be obtained using certain income-splitting techniques. For example, the attribution rules contained in subsections 74.1(1) and (2) of the ITA can serve to negate the tax advantages of transferring an income-generating property to a family member. As the respondent notes, similar measures can be traced back to the Income War Tax Act, 1917. Similarly, subsections 56(2) to (4) serve to regulate certain transactions by which income is diverted from an individual to the individual’s spouse.
 However, certain provisions in the ITA also encourage or facilitate some arrangements that arguably constitute income splitting. For example, under subsection 146(5.1), a taxpayer is entitled to a deduction in respect of a contribution to a spouse’s registered retirement savings plan. Similarly, by virtue of subsections 146.1(5) and (6), the income generated within a child’s registered education savings plan is generally payable in respect of (in the hands of either the child or the contributor). Additionally, a taxpayer can generally transfer capital property to his or her spouse on a tax-deferred basis under subsection 73(1) (although any income generated by the property or any resulting capital gain will generally be attributed back to the transferor pursuant to section 74.1 or 74.2).
 Another relevant contextual feature of the ITA relates to the practice of surplus-stripping and the provisions of the ITA that address it. There are multiple provisions in the ITA that can serve to eliminate the tax advantages of certain surplus stripping transactions. The appellants cite sections 84 and 84.1 of the ITA, which deem certain transactions, which might not be viewed as dividend transactions under corporate law, to give rise to dividends for the purposes of the ITA. For example, subsection 84(3) will generally deem a corporation to have paid a dividend where the corporation has redeemed, acquired or cancelled any of the shares of its capital stock and the amount paid by the corporation exceeds the paid‑up capital of the shares redeemed, acquired or cancelled.
 Similarly, section 84.1 aims at precluding the stripping of corporate surpluses where shares are sold in non-arm’s length transfers implemented through transaction steps similar to those under scrutiny in these appeals. In very general terms, section 84.1 prevents individuals from extracting corporate surpluses by, inter alia, triggering capital gains in respect of which a capital gains exemption is claimed under section 110.6. The amount of corporate surplus extracted directly or indirectly by a non-arm’s length party through the use of a capital gains exemption is deemed to be a dividend. For example, assuming that common shares of FHDM were qualified small business shares in that they were shares of a small business corporation as defined in the ITA, section 84.1 could have applied if the trust had sold common shares and the appellants had been entitled to claim, and did in fact claim a capital gains exemption under section 110.6 in respect of the capital gains allocated to them.
 Section 212.1 of the ITA operates in a similar fashion where a non-resident of Canada seeks to extract corporate surpluses by, inter alia, triggering directly or indirectly a capital gain that is, in most cases, exempt from tax by virtue of an applicable tax treaty.
 The fact that specific anti-avoidance provisions were enacted long before the introduction of section 120.4 leads me to infer that Parliament was well aware of the fact that taxpayers could arrange to distribute corporate surpluses in the form of taxable dividends or of capital gains subject to the application of those specific anti‑avoidance provisions. The fact that those provisions were not amended and that a specific rule was not included in section 120.4 to curtail well-known techniques leads me to infer that Parliament preferred simplicity over complexity when it enacted section 120.4. This militates against a finding that section 120.4 is indicative of a general policy in the ITA against surplus stripping implemented with the help of non-arm’s length parties.
 The appellants also cite section 15 of the ITA, which can deem certain capital transactions between a corporation and a shareholder (i.e., section 15 benefits) to be on account of income. For example, subsection 15(1.1) can require a shareholder to recognize an income inclusion equal to the fair market value of a stock dividend in circumstances where “it may reasonably be considered that one of the purposes of [the] payment [of a stock dividend] was to significantly alter the value of the interest of any specified shareholder of the corporation”. Certain section 15 benefits are specifically included in the definition of “split income” contained in subsection 120.4(1). I surmise that the respondent did not invoke this provision in the instant case because value was shifted from shares owned by the Trust to other shares owned by the Trust rather than from one taxpayer to another.
 The context of section 120.4 demonstrates that Parliament has implemented many measures within the framework of the ITA to reduce the effectiveness of surplus stripping. This contrasts with the respondent’s submission that Parliament did not foresee the use of capital gains transactions to strip corporate surpluses when enacting section 120.4. The existence of the above-mentioned specific measures suggests that, when enacting section 120.4, Parliament was well aware of the practice.
(i) 1999 Budget, Supplementary Information
 It is well established that budget materials are appropriate extrinsic materials to consider when engaging in a purposive interpretation of a provision of the ITA. In this case, there are certain extrinsic materials that are relevant. In 1999, the Department of Finance published a document entitled The Budget Plan 1999, Including Supplementary Information and Notices of Ways and Means Motions (February 16, 1999) (the “1999 BSI”), which contained the following commentary on the draft version of section 120.4:
In order to improve the fairness and integrity of the Canadian tax system, this budget proposes a targeted measure to discourage income splitting with minor children. The new measure constitutes a special tax, at the top marginal tax rate instead of the normal graduated rates, to be imposed on certain income of individuals age 17 or under.
Income that is not received as dividend income, partnership income or trust income is not subject to the new measure. Accordingly, income from employment or personal services of the minor will not be subject to this measure. Dividends received on any listed shares will not be subject to these rules, since the income flow is less susceptible to manipulation. Further, income from property acquired on the death of a parent of the individual will be exempt from the measure, as will income from any property inherited by individuals with disabilities who are eligible to claim the disability tax credit or by individuals who are in full time attendance at a post-secondary institution. Individuals who have no parent resident in Canada for tax purposes in the year will also be exempted from the application of the new tax.
The scope of this new measure is narrow; it targets those structures that are primarily put in place to facilitate income splitting with minors. The government will monitor the effectiveness of this targeted measure, and may take appropriate action if new income-splitting techniques develop.
 The respondent cites these comments from the 1999 BSI in support of the proposition that Parliament intended section 120.4 to prevent income splitting achieved through the use of forms of income that are susceptible to manipulation. The respondent relies on the Department of Finance’s reference to the fact that dividends on listed shares are to be excluded from the tax on split income because they are less susceptible to manipulation. However, the fact that it may have excluded certain types of income because they are less susceptible to manipulation does not mean that it intended to include all forms of income that are susceptible to manipulation. Moreover, the comments in the 1999 BSI suggest that, in enacting section 120.4, Parliament did not intend it to be a broad provision dealing with all types of income splitting with minors. The Department of Finance stated that the proposed measure would only apply to dividend income, partnership income or trust income, and emphasized that the provision is “narrow” and “targeted”.
(ii) Subsequent Amendments to Section 120.4
 As discussed above, subsequent amendments to a provision, may, in limited circumstances, be a relevant consideration when examining the policy underlying the former version of the provision.
 In 2011, section 120.4 was amended to add new subsections 120.4(4) and (5), which are applicable to dispositions occurring after March 21, 2011. In general terms, these new subsections operate to deem a specified individual who has received certain capital gains, either directly or through a trust, to have received non-eligible, taxable dividends for the purposes of the ITA (including subsection 120.4(2)). Subsection 120.4(4) generally applies where a specified individual realizes a capital gain on the disposition of shares of a private corporation to a person with whom the individual does not deal at arms length. Subsection 120.4(5) generally applies where a specified individual would be required to recognize trust income that can reasonably be considered to be attributable to a capital gain realized on the disposition of shares of a private corporation to a person with whom the specified individual does not deal at arms length. It would appear that, had the 2011 amendment to section 120.4 been in force when the transactions in issue were carried out, the Relevant Capital Gains would, by virtue of subsection 120.4(5), have given rise to tax under subsection 120.4(2).
 In conjunction with the proposal to amend section 120.4 by the addition of subsections 120.4(4) and (5), as described above, the Department of Finance released a document entitled Tax Measures: Supplementary Information and Notices of Ways and Means Motions (June 6, 2011). In that document, the Department of Finance made the following comments about the proposal (at page 287):
The tax on split income did not initially apply to capital gains because the planning techniques that were being used at the time did not rely on capital gains to split income with a minor. However, income-splitting techniques have been developed that use capital gains to avoid the tax on split income. These techniques involve capital gains being realized for the benefit of a minor on a disposition of shares of a corporation to a person who does not deal at arm’s length with the minor.
Budget 2011 proposes a targeted measure to maintain the integrity of the tax on split income regime. The measure will extend the tax on split income to capital gains realized by, or included in the income of, a minor from a disposition of shares of a corporation to a person who does not deal at arm’s length with the minor, if taxable dividends on the shares would have been subject to the tax on split income. Capital gains that are subject to this measure will be treated as dividends and, therefore, will not benefit from capital gains inclusion rates nor qualify for the lifetime capital gains exemption.
This measure will apply to capital gains realized on or after March 22, 2011. In addition, the government will continue to monitor the effectiveness of the tax on split income regime and will take appropriate action if new income-splitting techniques develop.
 These comments suggest that Parliament’s intention in amending section 120.4 was to broaden the scope of the provision. The Department of Finance notes that the measure will “extend” the tax on split income. There is no suggestion that the provision had as its purpose the prevention of all income splitting with minors. Indeed, the Department of Finance suggests that the amendment is itself a “targeted measure”.
 It is also noteworthy that the amendment extends the so-called kiddie tax to a broader subset of capital gains transactions. The amendment includes capital gains realized on the disposition of non-listed shares by specified individuals regardless whether the transactions include or not a corporate-surplus-stripping purpose or result such as that alleged in the instant case. In my opinion, a reasonable inference can be drawn that Parliament decided not to cover capital gains when the measure was first enacted, and chose to do so on a prospective basis only with respect to a narrow subset of capital gains transactions.
 Contrary to the submissions of the respondent, the manner in which the 2011 amendment was proposed and enacted does not suggest that Parliament was closing a loophole as described in Water’s Edge. In this case, Parliament moved to prevent a tax-planning strategy approximately ten years after that strategy was first outlined publicly. To paraphrase Nöel J.A.’s comments in Triad Gestco, when Parliament introduced section 120.4 it had to be aware that capital gains could be used to counter its operation.
 The respondent, in her written submissions, argues that the appellants succeeded in creating and triggering artificial capital gains by using high-low dividends, as follows:
61 In this case, the appellants undertook transactions designed to create and trigger a capital gain at their choosing and timing through the high-low stock dividend transaction.
62 The "amount" of a stock dividend is defined in subsection 248(1) as generally the amount by which the paid-up capital of the corporation that paid the dividend is increased by reason of the payment of the dividend. Likewise, the adjusted cost base of the shares issued on the stock dividend is defined in subsection 52(3) as the same amount. In this case, the stock dividend was purposely set low and the fair market value purposely set high such that the amount of the dividend to the Trust under section 82 was negligible and the created pregnant gain could be deferred to a time of the taxpayer’s choosing. The creation of the high-low stock dividend simply transferred the existing value in FHDM to the minor children.
63 The trust then sold the shares to the father who in turn sold the shares to a non-arm’s length corporation. The dispositions and the capital gains were spread over three years to secure the lowest tax rate and the redemption was of the corporation’s shares so as to access the section 112 dividend deduction. The minor children were allocated the funds through exchanged notes. Not too dissimilar transactions were also in play, but for a different result in Triad Gestco and two other recent "value shift" cases from the Federal Court of Appeal. Here, the transactions were designed to circumvent subsection 120.4 through the creation and manipulation of a capital gain.
 Counsel for the respondent suggests that the transactions in issue have the same circularity and artificiality as the transactions at issue in Triad Gestco and 1207192, and that the transactions in issue similarly did not reflect an increase in real economic power. With respect, this argument is a red herring.
 While the transactions in Triad Gestco and 1207192, like the transactions in issue in these appeals, relied upon high-low shares, the similarity appears to end there. In this case, a capital gain had accrued on the common shares. The appellants did not need to resort to the payment of a stock dividend to trigger a capital gain. The Trust could have realized the same capital gain by selling the appropriate number of common shares of FHDM to trigger the targeted capital gain and by allocating the resulting gain to the appellants. The stock dividend did not create a capital gain as contended by the respondent. All that the stock dividend achieved was to transfer part of the accrued gain on the common shares to the Class D Preferred Shares. I suspect that this was done in order to dispense with a valuation of the common shares, which would have been required had the transaction proceeded instead through a sale by the Trust of some of the common shares of FHDM. A valuation was not required for the Class D Preferred Shares because their redemption price was fixed and the residual value of FHDM accrues to its common shares. Whereas in Triad Gestco and 1207192 the taxpayers created artificial capital losses to shelter true economic capital gains, the appellants in this case paid tax on their capital gains. Moreover, the respondent’s concession with respect to the capital gain recognized by Steven for his 2005 taxation year is inconsistent with the position that the Relevant Capital Gains were artificial. There is no principled distinction in this regard between the Relevant Capital Gains and the capital gain recognized by Steven for his 2005 taxation year.
 For all of the reasons noted above, I am of the opinion that the transactions giving rise to the Relevant Capital Gains did not circumvent the application of section 120.4 in a manner that constituted abusive tax avoidance for the purposes of subsection 245(4). At the very least, the respondent did not establish that the transactions violated subsection 245(4). Therefore the appeals are allowed and the assessments are vacated.
Signed at Calgary, Alberta, this 1st day of May 2013.
"Robert J. Hogan"