Brelco Drilling Ltd. v. R., 99 DTC 5253, [1999] 3 CTC 95 (FCA)

Linden J.A. (Isaac C.J. concurring):

I. Introduction

This appeal deals with one of the most complex sections of the Income Tax Act, R.S.C. 1985, 5 Supp., C. 1, as amended (“ITA”). The issue is whether the Tax Court Judge erred in law when he held that it was unnecessary for purposes of section 55(2) of the ITA to consider the losses and other liabilities of foreign subsidiaries in determining the deemed “income earned or realized” pursuant to section 55 of the ITA.

II. Interpreting the Income Tax Act

The Supreme Court of Canada has on many occasions discussed the proper way to interpret the Income Tax Act, stating that the Court should take a contextual and purposive approach to statutory interpretation unless the meaning of the provision is clear and plain. In the seminal case of Stubart Investments Ltd. v. R.,! [1] Estey J. relied on the following passage from E. A. Driedger [2] :

Today there is only one principle or approach, namely, the words of an Act are to be read in their entire context and in their grammatical and ordinary sense harmoniously with the scheme of the Act, the object of the Act, and the intention of Parliament.

Despite occasional confusing and inconsistent statements in the jurisprudence, reading the words of the Act in their total context is still the proper approach to the ITA as long as the meaning is not “clear and plain”. [3] In Antosko, lacobucci J. wrote for a unanimous Court that, if the meaning is “clear and plain”, that governs, but if it is not, context and purpose must be investigated:

While it is true that the courts must view discrete sections of the Income Tax Act in light of the other provisions of the Act and of the purpose of the legisla- tion, and that they must analyze a given transaction in the context of economic and commercial reality, such techniques cannot alter the result where the words of the statute are clear and plain and where the legal and practical effect of the transaction is undisputed.^ [4]

Unfortunately, though, it is rarely easy to ascertain the plain meaning of a provision of the ITA. As Mr. Justice Cory wrote, in Canada Trustco Mortgage Corp. v. Port O'Call Hotel Inc. for the majority of the Supreme Court:

...I accept the following comments on the Antosko case in P.W. Hogg and J.E. Magee, Principles of Canadian Income Tax Law ... at 453-454:

It would introduce intolerable uncertainty into the Income Tax Act if clear language in a detailed provision of the Act were to be qualified by unexpressed exceptions derived from a court’s view of the object and purpose of the provision.... (The Antosko case) is simply a recognition that “object and purpose’’ can play only a limited role in the interpretation of a statute that is as precise and detailed as the Income Tax Act. When a provision is couched in specific language that admits of no doubt or ambiguity in its application to the facts, then the provision must be applied regardless of its object and purpose. Only when the statutory language admits of some doubt or ambiguity in its application to the facts is it useful to resort to the object and purpose of the provision.

[parai 5] Thus, when there is neither any doubt as to the meaning of the legislation nor any ambiguity in its application to the facts then the statutory provision must be applied regardless of its object or purpose. I recognize that agile legal minds could probably find an ambiguity in as simple a request as “close the door please’’ and most certainly in even the shortest and clearest of the ten commandments. However, the very history of this case with the clear differences of opinion expressed as between the trial judges and the Court of Appeal of Alberta indicates that for able and experienced legal minds, neither the meaning of the legislation nor its application to the facts is clear. It would therefore seem to be appropriate to consider the object and purpose of the legislation. Even if the ambiguity were not apparent, it is significant that in order to determine the clear and plain meaning of the statute it is always appropriate to consider the “scheme of the Act, the object of the Act, and the intention of Parliament’’. (Citation omitted)^ [5]

The Supreme Court’s direction is unambiguous: while courts are not to alter the wording chosen by Parliament, there are many sections in this Act

— which runs over 1000 pages of very small print — which are ambiguous or unclear. In those situations, courts can not abdicate their role as interpreters of the law and must examine the context and purpose of the legislation in order to give a fair meaning to the provision in question, if they can do so.

III. ITA section 55

Section 55, which contains some 4760 words, is ambiguous in many of its parts; to say that it has a “clear and plain meaning” is fanciful. The Tax Court Judge below described it in this manner:

In the case at bar, the Court was obliged to journey to areas where seasoned professionals recoil with more than feigned horror at even potential exposure to the legislative and regulatory morass respecting section 55 and the foreign affiliate Income Tax Regulations.^ [6]

In another case, [7] Finch, J., of British Columbia Supreme Court complained of the difficulty of section 55, saying:

It surpasses my imagination that anyone considers language such as this to be capable of an intelligent understanding, or that such language is thought to be capable of application to the events of real life, such as the sale of a business.

Despite these complexities, however, it is the task of this Court to do its utmost to give a sensible meaning to the provision. Section 55(2), the key provision in section 55, reads as follows:

55(2) Where a corporation resident in Canada has after April 21, 1980 received a taxable dividend in respect of which it is entitled to a deduction under subsection 112(1) or 138(6) as part of a transaction or event or a series of transactions or events (other than as part of a series of transactions or events that commenced before April 22, 1980), one of the purposes of which (or, in the case of a dividend under subsection 84(3), one of the results of which) was to effect a significant reduction in the portion of the capital gain that, but for the dividend, would have been realized on a disposition at fair market value of any share of capital stock immediately before the dividend and that could reasonably be considered to be attributable to anything other than income earned or realized by any corporation after 1971 and before the transaction or event or the commencement of the series of transactions or events referred to in paragraph (3)(a), notwithstanding any other section of this Act, the amount of the dividend (other than the portion thereof, if any, subject to tax under Part IV that is not refunded as a consequence of the payment of a dividend to a corporation where the payment is part of the series of transactions or event)

(a) shall be deemed not to be a dividend received by the corporation;

(b) where a corporation has disposed of the share, shall be deemed to be proceeds of disposition of the share except to the extent that it is otherwise included in computing such proceeds; and

(c) where a corporation has not disposed of the share, shall be deemed to be a gain of the corporation for the year in which the dividend was received from the disposition of a capital property.

55(2) Dans le cas où une société résidant au Canada a reçu un dividende imposable à l’égard duquel elle a droit à une déduction en vertu des paragraphes 112(1) ou (2) ou 138(6) dans le cadre d’une opération, d’un événement ou d’une série d’opérations ou d’événements dont l’un des objets (ou, dans le cas d’un dividende visé au paragraphe 84(3), dont l’un des résultats) a été de diminuer sensiblement la partie du gain en capital qui, sans le dividende, aurait été réalisée lors d’une disposition d’une action du capital-actions à la juste valeur marchande immédiatement avant le dividende et qu’il serait raisonnable de considérer comme étant attribuable à autre chose qu’un revenu gagné ou réalisé par une société après 1971 et avant le moment de détermination du revenu protégé quant à l’opération, à l’événement ou à la série, malgré tout autre article de la présente loi, le montant du dividende (à l’exclusion de la partie de celui-ci qui est assujettie à l’impôt en vertu de la partie IV qui n’est pas remboursé en raison du paiement d’un dividende à une société lorsqu’un tel paiement fait partie de la série):

(a) est réputé ne pas être un dividende reçu par la société;

(b) lorsqu'une société a disposé de l’action, est réputé être le produit de disposition de l’action, sauf dans la mesure où il est inclus par ailleurs dans le calcul de ce produit;

(c) lorsqu’une société n’a pas disposé de l’action, est réputé être un gain de la société pour l’année au cours de laquelle le dividende a été reçu de la disposition d’une immobilisation.

This provision was enacted by Parliament to counter “capital gains stripping”. [8] In the absence of this provision, corporations could, without attracting any tax, realize capital gains from related corporations by having those corporations issue dividends. The operation of provisions such as section 113 of the ITA avoided the tax payable on such dividends. [9] In the past, corporations engaged in capital gains stripping, effectively realizing capital gains held by affiliates without attracting tax. Rather than avoiding double taxation, this technique circumvented taxation on amounts that had not been taxed at all. One practitioners’ resource describes capital gains stripping with two examples:

For example, prior to the enactment of section 55(2), a corporation that desired to sell its shares in another corporation could first receive a tax-free dividend. It would then sell its shares at a price reduced by the amount of the dividend, thereby decreasing the capital gain which would otherwise have been realized had the sale been made without a dividend being declared. A seller might also avoid the realization of capital gains on other property it owned by transferring the property to the purchasing corporation using the subsection 85(1) election and receiving back preferred shares with a redemption price equal to the value of the transferred property but with a low paid-up capital. Most of the proceeds on a redemption of the shares would be deemed to be a tax-free dividend in the hands of the corporation. æ [10]

In order to prevent capital gains stripping, section 55(2) attributes certain unrealized capital gains to the corporation receiving an otherwise tax- free dividend. Specifically, 55(2) deems certain funds received as an otherwise tax-free dividend to be a capital gain. This Court has described the operation of section 55(2) as follows:

[I]f it can be shown that the entire amount of the dividend is attributable to or covered by safe income, then subsection 55(2) is not applicable. If, however, a portion of the dividend or capital gain is attributable to something other than

113(1) Where in a taxation year a corporation resident in Canada has received a dividend on a share owned by it of the capital stock of a foreign affiliate of the corporation, there may be deducted from the income for the year of the corporation for the purpose of computing its taxable income for the year, an amount equal to the total of

(a) an amount equal to such portion of the dividend as is prescribed to have been paid out of the exempt surplus, as defined by regulation (in this Part referred to as “exempt surplus”) of the affiliate,

(b) ...

(C) …

(d) ...

The purpose of section 113 is to avoid double taxation. Section 55 was enacted with section 113 in mind, because corporations were using section 113 to strip gains from foreign affiliates and avoid tax altogether.

safe income, then the entire amount received by the corporation is deemed not to be a di vidend J [11]

Thus in my view section 55(2) operates as follows:

(a) The taxpayer must be a corporation resident in Canada.

(b) The taxpayer must receive a taxable dividend for which certain deductions (ss. 112(1), 112(2), or 138(6)) apply.

(c) The dividend must be received as part of a transaction or series of transactions.

(d) One of the purposes of the transaction must be to effect a significant reduction in capital gains which would be realized if the share had been sold by the recipient corporation just before the dividend.

(e) That portion of the capital gain which can reasonably be attributable to anything other than “income earned or realized” is attributed to the corporation as capital gain. [12]

The key phrase from section 55(2) reads

Where a corporation resident in Canada has ... received a taxable dividend ... one of the purposes of which ... was to effect a significant reduction in the portion of the capital gain that, but for the dividend, would have been realized on a disposition at fair market value of any share of capital stock immediately before the dividend and that could reasonably be considered to be attributable to anything other than income earned or realized by any corporation ... (Emphasis added).

Due to the operation of this phrase, to avoid double taxation, section 55(2) does not apply to a dividend (or that portion of the dividend) which can reasonably be considered to be attributed to income earned by the corporation issuing the dividend. That income is referred to colloquially as “safe income.” - [13] The more “safe income” held by the corporation issuing the dividend, the larger the tax-free dividend which can be issued.

Revenue Canada has adopted the position that, in order for income to be “safe,” it must be income on hand. In Placer Dome, supra, this Court ex- pressly accepted Revenue Canada’s position that safe income must be income on hand. Robertson J.A. wrote for a unanimous panel that:

It is common ground that subsection 55(2) does not apply to dividends wholly attributable to income earned by Falconbridge after 1971 or what is referred to in tax circles as “safe income”. The amount that can be paid as a dividend without offending subsection 55(2) is colloquially referred to as a “safe dividend” and can be paid to the extent that there is safe income on hand immediately before the dividend is paid.... (Emphasis added) [14]

In Walnut Investment, supra, this Court expressly accepted Revenue Canada’s valuation of safe income. Robertson J.A. explained:

Subsection 55(2) applies if a dividend effects a significant reduction in a capital gain that could reasonably be considered to be attributable to anything other than safe income. Accordingly, assuming that the other requirements of that provision are satisfied, so long as the approach taken by the Minister in allocating safe income is reasonable, subsection 55 (2) should apply regardless of whether the method chosen by [the appellant] could also be considered reasonable. (Emphasis added.) IS [15]

To avoid the operation of section 55(2) the taxpayer bears the onus “to establish the inapplicability of section 55(2) of the Act.” [16] In other words, he must show that the dividends were paid out of “safe income,” which requires the calculation of funds on hand to pay the dividend. In this way, the matter is not unlike any other tax case: the Minister re-assesses the taxpayer, and in doing so sets out a series of factual assumptions. It falls to the taxpayer to rebut those assumptions on the facts, or to show that the Minister’s assumptions do not bear the legal consequences contended for.

To assist with the interpretation of section 55(2), Parliament enacted sections 55(3) to 55(5). Section 55(3)(a) exempts true intra-corporate dividends from the operation of section 55(2). Sections 55(3)(b), 55(3.1), and 55(3.2) set out the application of section 55(2) to so-called butterfly reorganizations. Thankfully, these complex provisions have little application to the case at bar. Section 55(4) is a further anti-avoidance provision. It reads as follows:

55(4) For the purposes of this section, where it can reasonably be considered that one of the main purposes of one or more transactions or events was to cause 2 or more persons to be related to each other or to cause a corporation to control another corporation, so that subsection (2) would, but for this subsection, not apply to a dividend, those persons shall be deemed not to be related to each other or the corporation shall be deemed not to control the other corporation, as the case may be.

55.(4) Pour l’application du présent article, lorsqu’il est raisonnable de considérer que l’un des principaux motifs d’événements ou d’opérations consiste à faire en sorte que des personnes deviennent liées entre elles ou qu’une société en contrôle une autre, de façon que le paragraphe (2) ne s’appliquerait pas, n’eût été le présent paragraphe, à un dividende, ce lien et ce contrôle sont réputés ne pas exister.

Thus, where corporations enter into transactions aimed at appearing to be (or at becoming) related corporations in order to avoid the operation of section 55(2), the corporations are deemed not to be related to each other. While this provision is not directly related to this case, it gives some idea of the breadth and flexibility which Parliament intended to give to the antiavoidance devices in section 55.

Section 55(5) is a provision which sets out some rules for determining safe income. Specifically, section 55(5)(d) sets out rules for calculating safe income from foreign affiliates. That provision reads:

55(5) For the purposes of this section, …

(d) the income earned or realized by a corporation for a period ending at a time when it was a foreign affiliate of another corporation shall be deemed to be the total of the amount, if any, that would have been deductible by that other corporation at that time by virtue of paragraph 113( 1 )(a) and the amount, if any, that would have been deductible by that other corporation at that time by virtue of paragraph 113( 1 )(b) if that other corporation

(i) owned all of the shares of the capital stock of the foreign affiliate immediately before that time,

(ii) had disposed at that time of all of the shares referred to in subparagraph (i) for proceeds of disposition equal to their fair market value at that time, and

(iii) had made an election under subsection 93(1) in respect of the full amount of the proceeds of disposition referred to in subparagraph (ii); ...

55.(5) Pour l’application du présent article: ...

(d) le revenu gagné ou réalisé par une société pour une période se terminant à un moment où elle était une société étrangère affiliée d’une autre société est réputé être le total de la somme qui aurait été déductible à ce moment par cette autre société en vertu de l’alinéa 113( 1 )a) et la somme qui aurait été déductible à ce moment par cette autre société en vertu de l’alinéa 113(1)b) si celle-ci, à la fois:

(i) était propriétaire de toutes les actions du capital-actions de la société étrangère affiliée immédiatement avant ce moment,

(ii) avait disposé à ce moment de toutes les actions visées au sous-alinéa (i) en contrepartie d’un produit de disposition égal à leur juste valeur marchande à ce moment,

(iii) avait fait un choix en vertu du paragraphe 93(1) relativement au montant global du produit de disposition visé au sous-alinéa (11); …

The respondent’s argument, reduced to its simplest form, is that 55(5)(d) sets out a complete code for the calculation of safe income as regards foreign affiliates.

IV. Facts

At the Tax Court, the parties filed an agreed statement of facts which the Tax Court Judge accepted and summarized in this manner:

[5] On November 29, 1989 the Appellant received a cash dividend of $32,000,000 from Tricil and included same in income. It applied subsection 55(2) to that dividend resulting in it being deemed to be proceeds of disposition. It reduced the resulting deemed capital gain by the amount of “safe income’’ calculated by it. It had losses from other operations for that taxation year, the ultimate position of the Appellant being that it had a non-capital loss for that year.

[6] On December 29, 1989 the Appellant sold all of its shares of Tricil to Laidlaw Inc.

[7] In 1995 the Minister, pursuant to the Appellant’s request, issued a Notice of Determination of Loss. It determined the loss to be less than the amount claimed by the Appellant, the difference relating directly to the amount of “safe income’’ which could be applied to reduce the amount of the capital gain.

[8] Each of Inc., A, B, C, D and E [affiliates of Tricil, Inc.] had an “exempt deficit” within the meaning of that expression for the purposes of the Act at the end of its taxation year ending prior to November 30, 1989. Each of B, C, D, and E incurred losses between its last prior taxation year and November 29, 1989.

[9] Each of F and G [affiliates of Tricil, Inc.] had an exempt surplus within the meaning of that expression for the purposes of the Act at the end of its taxation year ending prior to November 30, 1989.

[10] The Appellant computed its “safe income” as being $25,735,216 and deducted that sum from the amount of deemed capital gain under subsection 55(2). The Minister calculated the portion of the dividend from Tricil to Trimac that could reasonably be considered to be attributable to income earned or realized by Tricil and its subsidiaries after 1971 (hereinafter referred to as “Safe Income”) ... to be $23,149,721, a difference of $2,585,495. This resulted in an increase in the Appellant’s taxable capital gain of $1,723,672 and a corresponding reduction in its non-capital loss.

[11] The Minister prepared an Analysis of Safe Income Calculation of the Appellant which was attached as a schedule to the Agreed Statement of Facts. It shows, in respect of Tricil, the amounts of $9,026,833 and $2,845,896 totalling $12,776,622. The parties agree that the total of these two amounts should be $11,872,729. Therefore, one-half of the resulting $903,893 overstatement of “safe income” should reduce the Appellant’s share by $451,946. It was agreed that if the Appellant succeeds in this appeal, this calculation error will be rectified resulting in a reduction of the Appellant’s computation of “safe income”. If, however, the Appellant is unsuccessful, the calculation error will not be rectified because the result would be an increase in the amount determined by the Court to be payable by the Appellant.

[12] The Respondent’s computation reduces the Appellant’s “safe income” by the amount of the exempt deficits of Inc., A, B, C, D and E. The Appellant contends that the exempt deficits should not be “netted” with the exempt surpluses of F and G. The Respondent submits that all exempt deficits should be so “netted”. (Footnotes omitted.)! / [17]

The core facts in this case are therefore quite simple. In the taxation year, 1991, Brelco (formerly Trimac) undertook a transaction by which the unrealized gain in certain subsidiaries, notably including foreign affiliates, were consolidated in the parent company, Brelco. The affiliates in question had both income and debts (losses). ITA section 55(5)(d), a deeming provision which sets the income levels of foreign affiliates for purposes of determining safe income [under s. 55(2)], does not on its face take the losses of the foreign affiliate into account. The Minister nevertheless took the position that the foreign affiliate’s losses reduced the safe cash on hand with which the affiliate could declare and pay the dividend, and thus needed to be considered for purposes of ITA section 55(2). Brelco appealed to the Tax Court of Canada.

After reviewing the facts and the argument of the parties at length, the Tax Court Judge rejected the government’s argument that section 55(2) was satisfied only after the taxpayer determined all the income which could not be attributed to income earned or realized by the affiliate corporation. Specifically, he held that since section 55(5)(d) set out specifically what calculations determined the income earned or realized by the affiliate corpora- tion, academic or governmental speculation regarding what other factors might enter 55(5) calculations were unwarranted by the ITA. He held that the use of such factors to calculate safe income were an attempt to give meaning to a statute “which is sorely wanting,” [18] reasoning that such an attempt to construe the legislation would be tantamount to rewriting the legislation.

V. Submission of the Appellant

Revenue Canada argues that if a foreign affiliate owes or pays taxes, owes or pays dividends, or suffers operational losses, those costs must reduce “income earned or gained,” because they are funds which are lost, i.e., not on hand to pay any dividend. To include those amounts as income earned or realized over-values the cash on hand, thereby wrongly inflating the potential amount of the tax-free dividends.

The appellant further submits that section 55(2) requires the Tax Court to decide what portion of the unrealized capital gain “could reasonably be considered to be attributable to” the income earned or realized in Brelco’s foreign affiliates. The appellant argues that Parliament expressly drew the section in a broad way in order to ensure that the Tax Court did not engage in hair-splitting over what was and was not to be calculated as part of safe income.

The appellant also argues that when Parliament used the words “could reasonably be considered to be attributable to…,” it permitted the Minister to assess tax on a reasonable calculation of safe income. In this case, Revenue Canada’s subtraction of the foreign affiliates’ losses from their safe income 1s, in its view, reasonable.

VI. Submission of the Respondent

The respondent argues that if Parliament wanted to reduce the amount earned or realized by foreign affiliates to reflect tax, dividends, or losses, it would have done so expressly. In the eyes of the respondent, section 55(5) specifically deems what income calculation is to be used for purposes of section 55(2). In the respondent’s view, there is no statutory authority for the appellant’s arguments that the phrase “can reasonably be attributable” allows the Minister to make any reasonable calculation of safe income.

Further, the respondent suggests that the finding of the Tax Court Judge is really a finding that the losses of the foreign affiliates bore no relationship to Brelco’s capital gain in respect of its shares in the affiliate.

Finally, the respondent contends that in this case there is neither evidence nor assumption that the losses of foreign affiliates were related to the surpluses of foreign affiliates, or that those losses had any adverse impact on the earnings or price of Brelco’s shares. When an asset is being valued, losses are only calculated to the extent that they contribute to the deterioration of share value, it is said.

VII. Analysis

The Tax Court Judge held that, as there is no provision in the Act requiring that losses be taken into account in determining “income earned or realized” for purposes of section 55(2), such losses need not be taken into account. I understand his primary reason to be that Revenue Canada’s position is not warranted by a plain reading of 55(2). I cannot agree.

I consider that the operation of section 55(2) has been settled. As noted above, in two recent decisions, unanimous panels of this Court have accepted that “safe income” means “safe income on hand.” Those panels declared that safe income was a net calculation. In one of those cases, Walnut Investment, supra, a unanimous panel of this Court expressly accepted that so long as the approach taken by the Minister in allocating safe income is reasonable, subsection 55(2) should apply. There is no reason why this Court should depart from its jurisprudence. The reasoning of the Tax Court Judge, that there is no legal basis for requiring exempt deficits — or any other cost which reduces the safe income on hand — to be taken into account in determining “income earned or realized”, is therefore flawed.

Section 55(2) is not clear and unambiguous. Where the words of a provision are difficult to understand, as here, Courts are required to look to the context, the object and the purpose of that provision. This Court did so in Placer Dome, supra, and Walnut Investment, supra. It would be wrong to alter that view now. While the application of those words may be difficult in a given case, courts cannot throw up their hands in defeat. Courts must assess, on the facts of the case, the value of safe income on hand. It is a difficult task, but not an impossible one.

I am bolstered in these conclusions by the literature, which unanimously accepts that section 55(2) requires a calculation of safe income on hand, not exempt income generally. “Safe income on hand” refers to that portion of the safe income of a share which can reasonably be considered to contribute to the capital gain on that share. It is by definition a net calculation which begins with the deemed income in section 55(5), but which does not end there. In an article which persuasively argues that the government’s view of safe income on hand is too broad, one tax practitioner writes:

For example, income taxes, declared and unpaid dividends, and disallowed expenses, although not deductible in the computation of safe income, can not generally be considered to contribute to the gain on shares of the corporation because the amounts have either been laid out ... or set aside ... to fulfil an obligation. Similarly, according to Revenue Canada, expenses that are not deductible for tax purposes and losses realized in the holding period also reduce the safe income because such amounts can never be on hand to contribute to the gain on the shares of the corporation. I? [19]

The same author goes on to argue that losses of a subsidiary should only be calculated as part of safe income if the parent is liable for, e.g.,has guaran- teed, them. [20] While other authors argue for [21] or against [22] the government’s characterization of the factors which make up “safe income” for purposes of s. 55(2), all the academic and judicial commentary regarding the factors which contribute to safe income agree on this: section 55(2) requires a calculation of safe income on hand based on the facts of each case. I regret to say that such a determination was not carried out in this case.

This Court has explained that the object and purpose of section 55 is to prevent capital gains stripping. This purpose is not inconsistent with the goal of avoiding double taxation. On the contrary, it was enacted to prevent abuse of the very section which seeks to avoid double taxation. With a proper factual determination of safe income, a court can determine the amount which, on the facts of the case, represents safe income on hand. [23] In the absence of evidence regarding what factors can reasonably be considered to contribute to the capital gain element of a declared dividend, a corporation might subvert the specific anti-avoidance purpose of section 55 with impunity, or Revenue Canada might include inappropriate factors in the determination of safe income on hand. In this case, the Tax Court Judge did not undertake the task of deciding the amount of safe income on the facts of this case. With respect, it was an error not to do so. While it was correct for the Tax Court Judge to accept the agreed statement of facts, he was also under an obligation to determine the true value of safe income on hand given those facts.

I am not persuaded that section 55(5)(d), reproduced above, forms a complete code as regards the calculation of safe income on hand for dividends paid by foreign affiliates. There are two reasons for this. First, such a reading of section 55(5) divorces the law from its original purpose and would permit capital gains stripping — the very mischief which section 55 seeks to prevent — in any situation where capital gain exists in foreign affiliates.

Second, the language of the operative provision, section 55(2), is intended to cast its net more broadly than is contemplated by the respondent’s argument. Section 55(2) applies to

the portion of the capital gain that, but for the dividend, would have been realized on a disposition at fair market value of any share of capital stock immediately before the dividend and that could reasonably be considered to be attributable to anything other than income earned or realized by any corporation... (Emphasis added)

Section 55(2) applies to “any corporation.” Had Parliament meant “any corporation, excepting foreign affiliates,” it would have said so. Section 55(2) is plainly intended to stop all capital gains stripping, not merely capital gains stripping by related domestic corporations. Further, the section applies to capital gains that could reasonably be attributed to anything other than income earned or realized. This necessitates sorting through all the factors which are alleged to affect safe income on hand. Where a foreign affiliate is involved, the calculation set out in section 55(5)(d) represents the beginning of the process, deeming amounts deductible under 113( 1 )(a) and 113( 1 )(b) to be income earned or realized by a corporation. This deeming, however, does not end the matter. Section 55(2) requires a net calculation of all those amounts which could reasonably be attributed to anything other than income earned or realized by the foreign affiliate.

While I find that the Tax Court Judge erred, I cannot accept the appellant’s approach, i.e., that in this case all losses must be taken into account because they reduce the safe income of the intercorporate dividend. That reasoning is flawed inasmuch as it may not reflect the actual value of the safe dividends. In order to decide whether the calculation presented by Revenue Canada is reasonable, the Tax Court must hear evidence as to which factors do or do not reduce safe income on hand. The Tax Court must consider the Minister’s assumptions and decide whether they amount to a reasonable calculation of safe income. In this task, the parties will no doubt choose to call expert and other evidence regarding the reasonableness of Revenue Canada’s calculation of safe income. In this case, no such consideration of the evidence was undertaken. This Court can not properly decide what is or is not an appropriate calculation of safe income on hand without evidence as to the actual effect of the various factors proposed both by the Crown and the taxpayer.

I would therefore allow the appeal with costs to the appellant, set aside the decision below, and remit the matter back to the Tax Court of Canada with instructions to determine, on the facts of this case, the true value of “safe income on hand.” That way, the portion of the capital gain which can reasonably be attributed to anything other than income earned or realized will be determined, following which the appropriate amount of tax-free dividends can be calculated and the proper amount of deemed capital gain, if any, can be determined.

Sexton J.A. (dissenting):

I have read in draft the reasons of Linden J.A. and with great respect find that I cannot agree that this case should be remitted back to the Tax Court Judge. I do not believe that the Tax Court Judge erred in finding that as a matter of law exempt deficits of foreign affiliates should not be deducted from the exempt surpluses of other foreign affiliates so as to reduce the parent corporation’s safe income. In addition, I have concluded that the Minister has not pleaded the necessary assumptions of fact to establish that the exempt losses had any impact on the parent’s income.

Facts

On November 29, 1989, the respondent, Brelco Drilling Ltd., (“Brelco”), formerly Trimac Limited (“Trimac”), received a cash dividend of $32 million from Tricil Ltd. (“Tricil”) in which it owned 50% of the shares. Subsequently, Trimac sold its shares in Tricil to Laidlaw Inc.. Tricil, which was a Canadian company, was a 100% shareholder of the U.S. resident corporation Tricil Inc. (“Inc.”). Hence, Inc. was a foreign affiliate of Tricil for the purposes of paragraphs 55(5)(d) and C, D, E, F, and G. Corporations F and G had in the aggregate exempt surpluses of $6,049,488. These surpluses represent amounts which could be repatriated by way of dividends to the respondent on a tax free basis by virtue of paragraph 113( 1 )(a) of the Act. The exempt surplus is deemed to be the “income earned or realized” by these corporations as foreign affiliates by virtue of paragraph 55(5)(d). Corporations A, B, C, D, E and Inc. had in the aggregate operational losses of $5,118,119, of which the respondent’s share was $2,975,398.One of the purposes of the $32 million dividend was to significantly reduce the taxable portion of the respondent’s unrealized capital gain on its Tricil shares. The respondent accordingly applied subsection 55(2) of the Act to the dividends deeming them to be proceeds of disposition to the respondent. In its 1989 taxation year the respondent designated $25,735,216 of the dividends as separate dividends pursuant to paragraph 55(5)(f) on the basis that this was the portion of the unrealized capital gain on the respon- dent’s Tricil shares which could reasonably be considered to be attributable to income earned or realized. The Minister assessed tax to the respondent for the 1989 taxation year on the basis that the safe income on hand available to the respondent to reduce the deemed proceeds of disposition was only $23,149,721. The difference between the respondent’s calculation of safe income on hand and the Minister’s calculation arose as a result of the Minister offsetting the exempt deficits of corporations A, B, C, D, E and Inc. against the exempt surpluses of corporations F and G. The respondent has taken the position that there should not be any such offset.

Statutory Provisions

The statutory sections relevant to this appeal are set out below:

55.(2) Where a corporation resident in Canada has after April 21, 1980 received a taxable dividend in respect of which it is entitled to a deduction under subsection 112(1) or 138(6) as part of a transaction or event or a series of transactions or events (other than as part of a series of transactions or events that commenced before April 22, 1980), one of the purposes of which (or, in the case of a dividend under subsection 84(3), one of the results of which) was to effect a significant reduction in the portion of the capital gain that, but for the dividend, would have been realized on a disposition at fair market value of any share of capital stock immediately before the dividend and that could reasonably be considered to be attributable to anything other than income earned or realized by any corporation after 1971 and before the transaction or event or the commencement of the series of transactions or events referred to in paragraph (3)(a), notwithstanding any other section of this Act, the amount of the dividend (other than the portion thereof, if any, subject to tax under Part IV that is not refunded as a consequence of the payment of a dividend to a corporation where the payment is part of the series of transactions or event)

(a) shall be deemed not to be a dividend received by the corporation;

(b) where a corporation has disposed of the share, shall be deemed to be proceeds of disposition of the share except to the extent that it is otherwise included in computing such proceeds; and

(c) where a corporation has not disposed of the share, shall be deemed to be a gain of the corporation for the year in which the dividend was received from the disposition of a capital property.

55.(5) For the purposes of this section

(d) the income earned or realized by a corporation for a period ending at a time when it was a foreign affiliate of another corporation shall be deemed to be the aggregate of the amount, if any, that would have been deductible by that other corporation at that time by virtue of paragraph 113( 1 )(a) ... if that other corporation

(i) owned all of the shares of the capital stock of the foreign affiliate immediately before that time,

(ii) had disposed at that time of all of the shares referred to in subparagraph (i) for proceeds of disposition equal to their fair market value at that time, and

(iii) had made an election under subsection 93(1) in respect of the full amount of the proceeds of disposition referred to in subparagraph (11).

93.(1) For the purposes of this Act, where a corporation resident in Canada so elects, in prescribed manner and within the prescribed time, in respect of any share of the capital stock of the foreign affiliate of the corporation disposed of by it or by another foreign affiliate of the corporation,

(a) the amount (in this subsection referred to as the “elected amount”) designated by the corporation in its election not exceeding the proceeds of disposition of the share shall be deemed to have been a dividend received on the share from the affiliate by the disposing corporation or disposing affiliate, as the case may be, immediately before the disposition and not to have been proceeds of disposition.

113.(1) Where in a taxation year a corporation resident in Canada has received a dividend on a share owned by it of the capital stock of a foreign affiliate of the corporation, there may be deducted from the income for the year of the corporation for the purpose of computing its taxable income for the year, an amount equal to the total of

(a) an amount equal to such portion of the dividend as is prescribed to have been paid out of the exempt surplus, as defined by regulation...

Analysis

At trial, the parties agreed upon a set of facts they considered relevant and which they considered to be adequate to dispose of the case. It was not submitted to this court by either party that more evidence was necessary in order to decide the appeal nor was it suggested that the matter be remitted to the Tax Court. The issue before the court was precisely stated by the Tax Court Judge at page 12 of his reasons:

The Respondent’s computation reduces the Appellant’s “safe income” by the amount of the exempt deficits of Inc., A, B, C, D and E. The Appellant contends that the exempt deficits should not be “netted” with the exempt surpluses of F and G. The Respondent submits that all exempt deficits should be so “netted”.

I do not accept that the facts agreed upon by the parties were insufficient to settle the issue before the court.

In essence, the appellant argues that the profits of successful subsidiaries should be consolidated with the losses of other subsidiaries, based on the wording of subsection 55(2). Thus this appeal hinges on a question of statutory interpretation, specifically on the meaning of the following words found in subsection 55(2):

one of the purposes of which was to effect a significant reduction in the portion of the capital gain that, but for the dividend, would have been realized on a disposition at fair market value of any share of capital stock immediately before the dividend and that could reasonably be considered to be attributable to anything other than income earned or realized by any corporation after 1971. [emphasis added]

The appellant argued that the proper approach was to examine the capital gain involved and analyze what portion could be reasonably considered to be attributed to something other than income earned or realized. In the appellant’s view, the test is not simply what is “income earned or realized”. Rather, only the portion of the “income earned or realized” that is on hand serves to reduce the taxable capital gain. Thus amounts, such as losses in foreign affiliates, which no longer represent value in the corporation must be subtracted from the amount of income earned by other affiliates calculated under subsection 55(5).

The learned Tax Court Judge rejected the Minister’s approach and concluded that the exempt surplus of F and G should not be reduced by the exempt deficits of A, B, C, D, E and Inc. in the computation of income earned or realized after 1971. He found:

[51] I am instructed by paragraph 55(5)(d) as to what “the income earned or realized by a corporation” is deemed to be for the purposes of section 55. Subsection (2) of Section 55 provides that the portion of a dividend “that could reasonably be considered to be attributable to anything other than income earned or realized by any corporation after 1971” shall be deemed not to be a dividend but to be proceeds of disposition. The exempt surplus of F and G, foreign affiliates of the Appellant, is by virtue of the interaction of paragraph 55(5)(d), subsection 93(1) and subsection 113(1) “income earned or realized by any corporation” within the meaning of that term in subsection 55(2). There is no comparable legislative or regulatory instruction respecting exempt deficits. Accordingly, I have concluded that the exempt surplus of F and G should not be reduced by the exempt deficits of A, B, C, D, E and Inc. in the computation of “income earned or realized after 1971”.

I agree with the learned Tax Court Judge. There is no provision in the Act requiring exempt deficits to be taken into account in determining “income earned or realized”. The only statutory requirement is contained in paragraph 55(5)(d), which establishes through s. 113( 1 )(a) and s. 93(1) that the income earned of a foreign affiliate is deemed to be its exempt surplus.

In oral argument, counsel for the Minister provided hypothetical examples in which he assumed that the “fair market value” of a parent corporation would be calculated by valuing the subsidiaries as a group. He argued that to determine the fair market value of the parent, the losses contained in the subsidiaries had to be deducted from what would otherwise be paid by the purchaser. With respect, this is an erroneous approach. I agree that the safe income earned by the parent cannot exceed its fair market value. However, the fact that some of the subsidiaries might have losses could be simply ignored by a purchaser. A purchaser could wind up the subsidiaries with an exempt deficit and in the absence of a guarantee would not be bound to pay off those losses. Thus a more accurate assessment of “fair market value” would involve calculating the sum total of the assets contained in the various corporations. The final purchase price would reflect the “fair market value” of those assets. In the instant case corporations A, B, C, D, E and Inc. added no value to Tried, and thus a purchaser would not pay an additional amount for them. However, the losses in those separately incorporated subsidiaries, in the absence of a guarantee of payment of those losses, will not decrease the “fair market value” of the parent. Indeed, it is possible a purchaser might be willing to pay more for the losses in those subsidiaries if those losses were useful for tax purposes.The passage quoted by my colleague from the Canadian Tax Journal supports my position:

It is important to note that there is no clear statutory authority for Revenue Canada’s position regarding the application of losses to the safe-income calculation. I question whether it is fair to state that the losses of a subsidiary generally reduce the safe income on hand of the parent corporation. In many cases, it would seem more appropriate to state that safe income should be computed as an aggregation of positive amounts. Under this approach, safe income in hand of a parent corporation would not be reduced by the losses or deficit of a subsidiary, except to the extent that the parent has guaranteed the losses of the subsidiary or is otherwise obligated to fund those losses. For example, assume that a parent corporation has $1 million of safe income and its subsidiary has a $1 million deficit and nominal capital. If the parent has not guaranteed the losses of the subsidiary, any value assigned to the shares of the parent will be reasonably attributable to safe income. In these circumstances, the losses of the subsidiary should not be applied to reduce the parent’s safe income to zero. To do so would, in effect, suggest that the subsidiary’s value is less than zero. Although the subsidiary’s shares are worthless, in many cases, it would be incorrect to state that they have a negative value, especially where the parent could simply abandon the subsidiary before the sale of its own shares. [24]

I agree with the view expressed above. Absent a guarantee or actual payment by the parent of subsidiary losses, it is difficult to conceive of an example in which the losses of foreign affiliates would affect the safe income of a parent. Moreover, in this case, it is abundantly clear that the Minister has failed to demonstrate that the losses had any impact on the respondent’s safe income. Before turning to this point, I will address the issue of safe income and the rationale behind paragraph 55(5)(d).

The calculation of safe income

In reaching his decision, the Tax Court Judge found that the correct starting point for determining safe income was with the income earned or realized. In his words, the simplest way to determine “anything other than income earned or realized by any corporation after 1971” is to:

commence with a computation of “income earned or realized” because paragraph 55(5)(d) states specifically what that amount is deemed to be. There is no provision in Part LIX of the Regulations requiring exempt deficits to be taken into account in determining “income earned or realized”. The only statutory requirement is contained in paragraph 55(5)(d). The inclusion of that provision in the Act ends this matter. *** [25]

In my view this is a sensible way of applying a very difficult section. My colleague has argued that this approach is inconsistent with comments made in Placer Dome Inc. v. R. [26] and Nassau Walnut Investments Inc. v. R.^ [27] With respect, I do not think these cases are relevant to the issue at hand, as neither discussed the question of whether subsidiary losses could reduce the safe income on hand of a separately incorporated parent. In Placer Dome Inc., the taxpayer disputed that the purpose of the transaction was to reduce the capital gain. Thus the issue was whether subsection 55(2) applied at all and the debate focused on the meaning of a different part of the subsection than the part at issue in the case at bar. Similarly, Walnut, which examined the operation of 55(5)(f) and the designation of separate dividends, is not on point to the instant case. In that case there was no discussion of safe income on hand which is relevant to the present case. As has been broadly recognized, it is difficult to discern the intent of Parliament from the words employed in subsection 55(2). However, it seems to me that if it was intended that deficits from one subsidiary should be offset against exempt surpluses of another it would have been clearly provided in the legislation. It was not. Thus, the Tax Court Judge correctly relied on Johns-Manville Canada Inc. v. /?. [28] to conclude that subsection 55(2) exhibited “reasonable uncertainty resulting from lack of explicitness which should be resolved in favour of the taxpayer”. The conclusion that foreign affiliate surpluses should not be netted with foreign affiliate losses advances another aim of the act: namely to reduce double taxation. The following examines the issue of double taxation as it relates to section 55.

Rationale of including exempt surpluses as income earned

The reduction of the taxable capital gain by the exempt surplus of a foreign affiliate is based on the same rationale that underlies the tax free treatment accorded to intercorporate dividends. It is well accepted that the rationale behind the tax-free treatment of intercorporate dividends is that the corporation which earns the money in the first instance is subject to taxation and it is not equitable to expose the same income to taxation a second time when it is passed by way of dividend from the subsidiary corporation to the parent. [29] In section 113, Parliament extended this preferential tax treatment to dividends flowing from the exempt surplus of a foreign affiliate to its Canadian parent. Section 55 was designed to prevent a taxable capital gain from becoming a tax-free intercorporate dividend. Subsection 55(5) establishes the rules for calculating a taxpayer’s safe income. Of special note is paragraph 55(5)(f) which allows the taxpayer to designate a portion of a taxable dividend as a separate dividend. With paragraph 55(5)(f), Parliament has expressed its intent that section 55 should operate without effecting double taxation. In Administration Gilles Leclair Inc. c. R. [30] the following comments were made to that effect:

It appears to be assumed by all authors who have written on the subject that the purpose of paragraph 55(5)(f) of the Act is to avoid the double taxation of corporate income. That provision is integrated into the capital gains taxation scheme provided for by subsection 55(2) of the Act in the case of share redemptions, which makes it possible not to tax as a capital gain the portion attributable to income earned or realized after 1971. [31] *

Paragraph 55(5)(f) ensures that any portion of the dividend that is attributable to the income earned of the taxpayer will not be considered as proceeds of disposition. As stated above, it protects the taxpayer from double taxation.

One of the principal reasons Parliament chose in paragraph 55(5)(d) to deem the exempt surplus of a foreign affiliate as its income earned for purposes of subsection 55(2) was to minimize double taxation. The appellant’s position, which calls for the deduction of exempt losses of other foreign affiliates in the computation of safe income, is outside the language of section 55 and directly negates the objective of minimizing double tax. In the present case, it should be noted that the affiliates are located in the United States, which has signed the Canada-United States Income Tax Convention (1980) with Canada. The primary purpose of this treaty is to minimize double taxation. In explaining the rationale behind the Convention, lacobucci J. stated it “was deemed important, in order to promote international trade between Canada and the U.S., to spare ... corporations double taxation”. [32] In my view, allowing the taxpayer to treat the full exempt surplus as safe income on hand recognizes, as does s. 113, that any further tax on the exempt surplus amounts to double taxation. Thus s. 55(5)(d) compliments s. 113 in ensuring that double taxation of foreign affiliate income will be avoided. I am convinced that the exempt surplus should not be offset by any exempt deficits for another reason. As separately incorporated companies, it is clear that each foreign affiliate in the United States is subject to income tax on its own earnings. The argument that one foreign affiliate could reduce its tax liability by deducting the losses of its sister affiliate would be rejected out of hand, as it would violate the elementary principle that each corporation is a separate entity and should be treated that way for tax purposes. Thus, it would be inconsistent and unfair to the parent corporation to subject it to tax on monies which it receives from a foreign affiliate when those monies have already been subject to tax in the hands of the foreign affiliate. In sum, my finding that foreign affiliate surpluses should not be netted with foreign affiliate losses is based on my interpretation of section 55 of the Income Tax Act. This interpretation is based on sound policy, as it promotes an important object of the Act by minimizing double taxation. I turn now to the issue of the deficiencies in the Minister’s pleadings.

Deficiencies in the Minister’s Pleadings

In reassessing the taxpayer, the Minister must set out a series of factual assumptions. The taxpayer is then required to disprove them. When analyzing the respective roles of the parties upon reassessment, the Tax Court Judge agreed with the Respondent that the case law establishes the following three principles. First, the Minister must assume all facts necessary to sustain the reassessment. Second, the appellant is not required to lead evidence in respect of facts not assumed. Third, references in the Reply to the Notice of Appeal to a statutory provision or conclusions of law are not allegations of fact or a pleading of fact. Based on these principles, I have come to the conclusion that the Minister has not pleaded the facts necessary to sustain the reassessment. In particular, the Minister did not plead any facts to establish that the losses in the foreign affiliates impacted the safe income of the parent. Further, there is no assumption or evidence explaining how the losses in the foreign affiliates arose, how the losses were funded or how those losses affected the safe income of the respondent. At paragraph 43, the Tax Court Judge noted that “certain assumptions of fact basic to the reassessment were not contained in the reply”. He then found “it seems unreasonable that the Respondent not be able to advance its case without having pleaded assumptions of fact to support the application of a provision to which the Appellant submitted itself.” I agree with the Tax Court Judge that it was not open to the taxpayer to contest the application of the section after choosing to submit to it. However, the Minister must still plead the factual assumptions to sustain the reassessment. In this case the Minister failed to connect the losses of the foreign affiliates with the income of the parent. Given this conclusion, I believe the proper course to follow is to reject the reassessment. I simply cannot agree that the matter should be remitted to the Tax Court. The case of del Valle v. Minister of National Revenue [33] ^, supports my view. In that case Sarchuk T.C.J. stated:

It is the appellant’s submission that there is nothing contained in the pleadings, nor for that matter in any of the material submitted by the respondent to the taxpayer, to indicate the facts essential to the validity of the assessments and that therefore the appellant has no case to meet. I did not take this to mean that counsel was suggesting that the Court should, on that basis alone, allow the appeal but that if the Court were to agree with this submission the respondent would have to bear the onus of proof with respect to the facts sought to be proven or any law applicable so as to support the reassessments under appeal.

It is settled law that a taxpayer is entitled to know the assumptions made by the respondent at the time of assessment particularly so because the onus is upon him to demonstrate that the basic assumptions of fact upon which the taxation rested are not capable of supporting the assessment. (See Johnston v. M.N.R., 3 D.T.C. 1182).

I believe this is the approach which should be followed in the case at bar. The Respondent has failed to allege as a fact an ingredient essential to the validity of the reassessment. There is no onus on the Appellant to disprove a phantom or non-existent fact or an assumption not made by the respondent.^ [34]

Here, the Minister did not plead the necessary assumptions or facts relating the foreign affiliate losses to the safe income of the parent. By failing to demonstrate how the exempt deficits of a foreign affiliate could impact the safe income of a separately incorporated parent, he is missing “an ingredient essential to the validity of the reassessment”. There is no evidence in the record and no suggestion that the losses or deficits of the subsidiaries A, B, C, D, E and Inc. were guaranteed by the respondent. In short, it was open to the Minister to plead the necessary facts at trial. He chose not to do so. In my view, it would be unfair to the taxpayer to now give the Minister a second opportunity to establish the validity of the reassessment.

Conclusion

In conclusion, section 55 of the Income Tax Act does not expressly require consolidation of the losses of one foreign subsidiary with the exempt surpluses of other foreign subsidiaries. The appellant’s interpretation of section 55 is not supported by the words of the section and appears to be contrary to the object and spirit of the Act. I would therefore dismiss the appeal with costs.

Appeal allowed.