Canada: CIBC v. The King, 2023 FCA 91.

Introduction

On May 4, 2023, the Federal Court of Appeal ("FCA") released its judgement¹ dismissing an appeal by Canadian Imperial Bank of Commerce ("CIBC") from a Tax Court of Canada ("TCC") decision² regarding the interaction of subsections 39(2)³ and 40(3.6) of the Income Tax Act (Canada)⁴ (the "ITA")⁵ in the context of a loss realized by CIBC in September, 2007 on the redemption of U.S. dollar-denominated shares it held in CIBC Delaware Holdings Inc. ("DHI"). The loss was solely attributable to the fluctuation in the value of the U.S. dollar ("US$") relative to the Canadian dollar ("C$") over the period in which CIBC held the shares of DHI.

While CIBC FCA concerned a historic version of subsection 39(2), the case raises an issue of potential significance to Canadian capital markets. As the Supreme Court of Canada has observed: "Canadian capital markets are of relatively modest size, and it is common for corporations requiring significant amounts of capital to look elsewhere in the world for financing."⁶ Canada's economy relies on access to foreign capital,⁷ and therefore the tax rules governing the treatment of foreign exchange gains and losses to a Canadian corporate issuer of debt in foreign markets must be clear.

The FCA's decision may arguably be viewed as introducing uncertainty Fregarding the application of both the historic version of subsection 39(2) and Amended Subsection 39(2). Specifically, the FCA held that subsection 39(2) cannot apply in and of itself to determine the amount of gains made or losses sustained because of foreign currency fluctuation on the disposition of shares. Such determination must first be made under separate provisions of the ITA. On its face, the decision is not strictly confined to the application of historic subsection 39(2) in the context of a disposition of property. Amended Subsection 39(2), which is limited to the settlement of foreign currency denominated debt, still contains the key words "made a gain or sustained a loss" that were considered in the FCA in interpreting historic subsection 39(2). However, the approach of first determining the amount of a gain made or loss sustained under a different provision of the ITA (i.e., section 40) before applying Amended Subsection 39(2) does not work on the settlement of debt. The settlement of debt is not a disposition of property by the issuer. The FCA did not address this distinction and, as a result, the decision may arguably be seen as introducing uncertainty regarding the recognition of foreign exchange gains and losses to Canadian issuers of foreign currency denominated debt.

The main issue in CIBC FCA was whether subsection 39(2) applied to deem CIBC's foreign currency related loss to be a capital loss from the disposition of foreign currency before the application of paragraph 40(3.6)(a), which would deem the loss from the redemption of the shares by DHI to be nil. CIBC argued that, consistent with the FCA's decision in Bank of Montreal,⁸ subsection 39(2) applied to deem the loss to be a capital loss from the disposition of foreign currency, and, therefore, there was no loss from the redemption of the shares to which 40(3.6) could apply.

In BMO, the FCA held that subsection 39(2) applied before subsection 112(3.1)⁹ in the context of a loss realized on the disposition in 2010 of US$-denominated common shares of a Nova Scotia unlimited liability company ("NSULC") by a limited partnership of which BMO was a limited partner. The Minister had reassessed Bank of Montreal ("BMO") under the general anti-avoidance rule ("GAAR") on the basis that BMO had avoided the application of subsection 112(3.1) by creating a class of preferred shares of NSULC on which dividends were paid. In dismissing the Crown's appeal, the FCA affirmed that subsection 39(2) could apply to foreign exchange losses arising from a disposition of property and held that subsection 39(2) applied to deem BMO's loss to be from a disposition of foreign currency. As such, there was no loss from the disposition of the shares to which subsection 112(3.1) could apply.

Seemingly at odds with the outcome in BMO, the TCC in CIBC TCC found that subsection 40(3.6) applied in priority to subsection 39(2) to deem CIBC's loss from the disposition of the shares of DHI to be nil. As such, there was no scope for the application of subsection 39(2). In CIBC FCA, the FCA dismissed CIBC's appeal, holding that the TCC did not err in its decision. Webb J.A. delivered the reasons for judgement on behalf of the FCA.

The decision in CIBC FCA raises significant interpretive issues involving the foreign currency fluctuation regime in subsection 39(2) and its interaction with other provisions in Subdivision c of Division B ("Subdivision C") of Part I of the ITA among others. One interpretative issue concerns the relationship between the different stop-loss rules (subsection 112(3.1) in BMO and subsection 40(3.6) in CIBC FCA) and subsection 39(2). A second interpretive issue concerns the words "sustained a loss" in subsection 39(2). In CIBC FCA the words were interpreted as requiring a loss to first be determined under other provisions of Subdivision C, namely section 40. Such interpretation implies that subsection 39(2) does not operate as a stand-alone provision and that other provisions that compute a taxpayer's gain or loss under section 40 take precedence.

We discuss and consider aspects of the FCA's reasoning in CIBC FCA in the context of the interpretive issues noted above. We start with a discussion of the relevant judicial history in the BMO and CIBC cases.

Judicial History: BMO and CIBC

BMO

BMO concerned the operation of subsection 39(2) in combination with subsection 112(3.1). The facts of the case are that BMO implemented an equity financing structure to fund certain U.S. subsidiaries. BMO became a partner in a Nevada limited partnership ("NLP") that acquired US$-denominated common shares and preferred shares of NSULC.

In order to hedge against certain currency fluctuation risks, the structure was organized such that dividends were paid on the preferred shares. The case was brought under the GAAR, with a key issue being whether the use of the preferred shares created a tax benefit. BMO argued, successfully at both the TCC and the FCA that there was no benefit on the basis that due to the operation of subsection 39(2), subsection 112(3.1) did not apply, such that the use of preferred shares to avoid subsection 112(3.1) was irrelevant and produced no tax benefit.

The TCC decision was concerned largely with the interpretation of subsection 39(2) and, in particular, whether the provision should be interpreted as applying exclusively to gains or losses arising on settlement of obligations on capital account or, more broadly, as applying to gains or losses arising from dispositions of any kind of property as well as on the settlement of obligations on capital account. The TCC found in favour of the latter interpretation (referred to as the "broad interpretation").

On appeal, the FCA upheld the TCC's decision. The FCA reviewed the text, context and purpose of subsection 39(2) and, drawing insight from the notwithstanding clause in subsection 39(2) and technical notes related to amendments to certain provisions in the foreign affiliate property income ("FAPI") regime (namely paragraphs 95(2)(f.12), 95(2)(f.15) and 95(2)(g.02)), concluded that the broad interpretation of subsection 39(2) applied.

In concluding that subsection 39(2) applied to deem the loss to be a capital loss from the disposition of foreign currency (and, therefore, not a loss from the disposition of the common shares to which subsection 112(3.1) could apply), the FCA stated:

The foregoing paragraph (the "BMO Paragraph") and, in particular, the comment that "[t]he gain or loss arising as a result of a disposition of a particular property was (and still is) determined under subsection 40(1) of the Act" (the "BMO Statement") became a significant point in CIBC TCC.

CIBC TCC

The facts in CIBC TCC are that, in November 2006, CIBC subscribed for 1000 Class B shares in DHI for a subscription price of US$1,000,000,000, which at the time was equivalent to C$1,130,000,000. In September 2007, DHI redeemed the 1,000 Class B shares held by CIBC for US$1,000,000,000, which at the time was equivalent to C$1,003,600,000. Following the redemption of the 1,000 Class B shares, CIBC continued to hold shares in DHI. As a result of the fluctuation of the value of US$ relative to C$, CIBC realized a capital loss of C$126,400,000 on the share redemption.

The CRA denied the loss on the basis that in order to determine a gain one must begin with subsection 40(1), which in turn requires a consideration of subsection 40(3.6). Paragraph 40(3.6)(a) would deem the loss realized on the share redemption to be nil and paragraph 40(3.6)(b) would, in computing CIBC's adjusted cost base of its remaining shares in DHI, add the amount of that loss to the adjusted cost base (determined without reference to paragraph 40(2)(g) and subsection 40(3.6)). The proceeding in the TCC was made by joint application of the parties for the determination of a question under section 58 of the Tax Court of Canada Rules. The question posed was whether paragraph 40(3.6)(a) of the ITA applies to deem CIBC's loss from the disposition of Class B shares of DHI to be nil.

The TCC began its analysis focussing on the BMO Statement. The TCC rejected CIBC's argument that the BMO Statement is obiter dicta. The TCC considered the BMO Paragraph an essential part of the reasoning of BMO, addressing the interaction between subsections 39(2) and 40(1).¹¹ CIBC posited that the BMO Statement was wrong at law. The TCC agreed with this position. In this regard, the TCC considered the text and context of the relevant provisions, namely subsections 39(1), 39(2), 39(3) and 40(1) at the time Subdivision C was introduced in 1972. The TCC stated:

The TCC went on to state that:

The TCC essentially agreed with CIBC's position that subsection 39(2) operated as a stand-alone provision that when applicable to a disposition of property replaces subsection 39(1) and the rules in Subdivision C. In this regard, "whether a gain is made, or a loss is sustained, for the purposes of subsection 39(2) is determined by reference to the commonly understood meaning of the words "gain" and "loss" and not by reference to the computational rules elsewhere in Subdivision c".¹⁴ Notwithstanding the above view of the TCC that the BMO Statement is incorrect, the TCC considered itself bound to follow the BMO Statement because it was part of the reasoning in BMO.¹⁵

CIBC FCA

The FCA upheld the TCC's decision. The FCA held that one must recognize that the ITA uses the terms gain/loss, capital gain/capital loss, taxable capital gain/allowable capital loss in distinct ways. The FCA cited the BMO Paragraph for the proposition that there was no conflict between subsections 39(2) and 40(1).¹⁶ In this regard, the FCA went on to state:

From this, the FCA concluded that a taxpayer could only have sustained a loss if the loss had been determined under another provision of the ITA because there was no formula in subsection 39(2) to determine the amount of the loss.¹⁸

In reaching the opposite conclusion to BMO, the FCA distinguished between the relevant stop-loss provisions at issue, namely subsection 112(3.1) in BMO and subsection 40(3.6) in CIBC FCA. The FCA correctly noted that subsection 112(3.1), as it applies to a corporate taxpayer, required the taxpayer to reduce its share of a loss allocated to it by a partnership (of which the taxpayer is a partner) by the amount of certain dividends received by the taxpayer that were deductible in computing the taxpayer's taxable income. The FCA cited a portion of the preamble to subsection 112(3.1) as follows:

The FCA interpreted the words "the loss determined without reference to this subsection" as meaning that the loss from the disposition of shares is first determined as if subsection 112(3.1) did not exist. The subsection only applied after the application of any other provision of the ITA that could change the amount of the loss or the source of the loss being from the disposition of shares.²⁰ According to the FCA, this was the case with respect to subsection 39(2), which in BMO, deemed the loss realized by BMO to be a capital loss from the disposition of foreign currency. Therefore, BMO no longer had a loss from the disposition of shares such that there was no loss to which subsection 112(3.1) could apply. ^[21]

In contrast, the FCA noted that there is no equivalent wording in paragraph 40(3.6)(a) to indicate that the loss (that is deemed to be nil) is to be determined without reference to subsection 40(3.6). While the FCA acknowledged that similar words appear in paragraph 40(3.6)(b), it noted that that paragraph only applies to the computation of the taxpayer's adjusted cost base in the remaining shares of the corporation held by the taxpayer, stating that "[i]t simply ensures that the amount added to the adjusted cost base of those other shares held by the taxpayer is the amount of the loss before such loss is deemed nil...".²² According to the FCA:

In essence, the FCA concluded that the presence of the above-noted words in the preamble of subsection 112(3.1) mandated that all of the provisions of Division B of the ITA, which includes subsections 39(2) and 40(1), be applied to determine the loss from the disposition of shares prior to any application of subsection 112(3.1). In contrast, the FCA interpreted the absence of these words in subsection 40(3.6) to mean that the provisions of section 40 were to be applied to determine the amount of the loss before subsection 39(2) could be applied.

In CIBC FCA the reasoning in respect of the stop-loss rules focuses on the fact that the phrase "the share of the loss determined without reference to this subsection..." is part of the loss reduction formula in subsection 112(3.1) but is not part of paragraph 40(3.6)(a). This interpretive issue in CIBC FCA is well canvassed and considered by other commentators.²⁴ Here, we focus on the issue of the operation of subsection 39(2) as a standalone provision.

Subsection 39(2) as a Standalone Provision

The notion that the amount of gains made or losses sustained that are attributable to foreign currency fluctuation must first be determined by formula under another provision of the ITA seems at odds with the historical application of subsection 39(2). Until BMO, the issue of whether subsection 39(2) could apply to a disposition of shares had not been directly considered by the Courts.

Traditional Application of Subsection 39(2) to Settlement of Foreign Currency Obligations

Historically, subsection 39(2) had been applied to foreign currency gains made or losses sustained on the settlement of capital obligations or liabilities. For example, in MacMillan Bloedel,²⁵ the corporate taxpayer, MacMillan Bloedel, had issued two series of US$-denominated preferred shares. The preferred shares were redeemable and retractable for their US$ issue price. As a result of an appreciation in the US$ relative to the C$ following the date of issuance, MacMillan Bloedel paid more to redeem the preferred shares than it received in subscription proceeds and claimed a capital loss under subsection 39(2). The Court held that MacMillan Bloedel was entitled to claim a capital loss under subsection 39(2) because it had sustained a loss by reason of a currency fluctuation. In reaching this conclusion, the Court likened MacMillan Bloedel's obligations under its preferred shares to those of a debtor with respect to payments owed under a loan.

The settlement of a capital obligation by a taxpayer is generally not considered to be a disposition of property and, therefore, the amount of any gain or loss is not computable or recognizable under the regime in subsections 39(1) and 40(1).²⁶ These subsections set out the ordinary way to determine a taxpayer's capital gain or capital loss from the disposition of a property. Paragraphs 39(1)(a) and (b) define a taxpayer's capital gain and capital loss for a taxation year from "the disposition of any property". Essentially, a taxpayer's capital gain from the disposition of property is the taxpayer's gain from the disposition determined under Subdivision C that would not be included in the taxpayer's income but for the application of certain rules in section 3, and a taxpayer's capital loss for a taxation year from the disposition of property is the taxpayer's loss from the disposition determined under Subdivision C that would not be deductible in computing the taxpayer's income but for the application of certain rules in section 3. Subsection 40(1) sets out the general rule for determining the amount of a taxpayer's gain or loss from the disposition of property, subject to the opening words "[e]xcept as otherwise expressly provided in this Part".

Given that the above regime only applies to dispositions of property, the amount of any gain or loss arising on settlement of a foreign currency-denominated debt attributable to fluctuations in the value of the foreign currency would not be computable under such regime. The interpretation given to subsection 39(2) by the FCA that requires the amount of gains made or losses sustained (by virtue of the fluctuation in the value of foreign currency) from the disposition of property to be first determined under the rules in section 40 seems inconsistent with the historical application of subsection 39(2) such as in MacMillan Bloedel. Such foreign currency gains or losses on settlement of a liability are not computed under any other provision of the ITA because they are not gains or losses from the disposition of property.

Notwithstanding Subsection 39(1)

The FCA's interpretation also appears to not give effect to the express words in subsection 39(2) beginning with "Notwithstanding subsection (1)". Parliament chooses its words carefully and each word of a provision must be given meaning.²⁷These words support the interpretation of subsection 39(2) as a standalone provision independent of the regime in subsections 39(1) and 40(1). In essence, subsection 39(2) ousts the ordinary way to determine a taxpayer's capital gain or capital loss from the disposition of a property, as described above, where the gain or loss is attributable to foreign currency fluctuation. In such case, subsection 39(2) applies to deem the gain or loss to be a capital gain or capital loss from the notional disposition of the foreign currency, arguably displacing the application of subsections 39(1) and 40(1).

As noted in CIBC TCC, the concept of making a gain or sustaining a loss was not entirely new to the ITA pre-1972. In Consolidated Glass Limited,²⁸ Rand J. (writing for Kellock and Fauteux, JJ.) stated:

The above words were used in paragraph 73A(1)(a) of the 1950 version of the ITA, well before the introduction of capital gains taxation in 1972 and the computational rules set out in subsections 39(1) and 40(1).

Thus, for purposes of subsection 39(2), the amount of a gain made or a loss sustained on a disposition of property by virtue of foreign fluctuations appears capable of determination under common law principles outside of subsection 40(1).³⁰ Such interpretation is consistent with the historical application of subsection 39(2) operating as a standalone provision with respect to gains or losses on the settlement of foreign currency-denominated debt and gives meaning to the words "notwithstanding subsection [39](1)".

Amended Subsection 39(2)

Subsection 39(2) was amended by the 2013 Amendments. The Department of Finance (Canada) explained the amendments as follows:

Given the changes, prima facie, the issue considered in BMO and CIBC FCA is no longer directly relevant given that Amended Subsection 39(2) no longer applies to foreign exchange gains and losses in respect of dispositions of property. However, despite the changes, certain key words contained in subsection 39(2) are carried over to Amended Subsection 39(2), the relevant parts of which read:

The words "made a gain or sustained a loss" were contained in old subsection 39(2). As noted above, the words sustained a loss were interpreted in CIBC FCA as requiring that, to sustain a loss for purposes of subsection 39(2), the loss must first be determined by formula under another provision of the ITA since subsection 39(2) contains no such formula.³²The interpretation derived from the TCC decision in CIBC TCC in which Owen J. considered the Court bound by the BMO Statement.

Interpretive Issues

It is difficult to reconcile the FCA's interpretation of these words for purposes of subsection 39(2) in the context of Amended Subsection 39(2). The issue is that CIBC FCA tells us that subsection 39(2) requires another provision to compute a gain or loss (i.e., section 40) and, on its face, the decision is not strictly confined to its facts. It is clear that Amended Subsection 39(2) is intended to only apply to foreign exchange gains or losses on the settlement of debt obligations of a taxpayer. It is well settled that the settlement of a debt obligation is not a disposition of property. Accordingly, the determination of a gain or loss under computational rules in subsection 40(1) is inapplicable to the settlement of a debt obligation because such rules only apply to determine a gain or loss on the disposition of a property.

The question is how does one determine a gain or loss for the purposes of Amended Subsection 39(2), taking into account that (i) CIBC FCA has told us that the amount of gains made or losses sustained cannot be determined under subsection 39(2) on a standalone basis, and (ii) gains and losses computed under section 40 only apply to dispositions of property. Further guidance and clarity from the Courts in answering this question – e.g., that CIBC FCA does not limit Amended Subsection 39(2) from operating on a standalone basis based on common law principles – would be highly welcome.

Conclusion

CIBC FCA raises important interpretative issues regarding the interaction of complex rules of the ITA and whether foreign currency regime in subsection 39(2) (and Amended Subsection 39(2)) is intended to operate on a standalone basis. The above discussion is intended to elicit thinking and debate on the issues and the reasoning of the Courts' decisions in CIBC.

Footnotes

1. Canadian Imperial Bank of Commerce v. The King, 2023 FCA 91. [CIBC FCA]

2. Canadian Imperial Bank of Commerce v. The Queen, 2021 TCC 71. [CIBC TCC]

3. Subsection 39(2) was amended in 2013 by 2013, c. 34, s. 59, applicable to gains made and losses sustained in taxation years that begin after August 19, 2011 (the "2013 amendments"). In this article, a reference to subsection 39(2) refers to the subsection prior to the 2013 amendments. We refer to subsection 39(2) post-2013 amendments as "Amended Subsection 39(2)".

4. R.S.C. 1985, c. 1 (5^th Supp.)

5. Unless otherwise indicated, all statutory references are to the ITA.

6. Imperial Oil Ltd. v. R., 2006 SCC 46, para 70 (dissent).

7. See for example, Statistics Canada's reporting on the Canadian debt denominated in foreign currency:

"Overall, foreign investors acquired $143.7 billion of Canadian securities in 2022. Foreign investment in corporate debt securities increased substantially in 2022 to reach unprecedented levels. New corporate bonds denominated in foreign currencies in the first half of 2022 led the investment activity." https://www150.statcan.gc.ca/n1/daily-quotidien/230227/dq230227a-eng.htm

Similarly, see the House of Commons comment on the reliance of the Canadian economy on trade in US dollars. "Given Canada's small size relative to the United States, which is the country's largest trading partner, Canada is generally a "price taker." According to Industry Canada, almost 80.0% of the value of Canada's manufacturing exports are destined for the United States and, thus, are traded in U.S. dollars. " https://www.ourcommons.ca/DocumentViewer/en/41-2/FINA/report-12/page-54

8. The Queen v. Bank of Montreal, 2020 FCA 82. [BMO]

9. See Elie S. Roth and Ryan Wolfe, "CIBC v Canada: A Matter of Precedence", International Tax Highlights, Volume 2,, Number 3, August 2023. The authors note that both the BMO and CIBC decisions turned on whether former subsection 39(2) was to be applied before or after the application of certain stop-loss rules.

10. BMO at para. 41.

11. CIBC TCC at para. 23.

12. CIBC TCC at para. 42-44

13. CIBC TCC at para. 48.

14. CIBC TCC at para. 55.

15. CIBC TCC at para. 24.

16. CIBC FCA at para. 27-28.

17. CIBC FCA at para. 29.

18. CIBC FCA at para. 31.

19. CIBC FCA at para. 36.

20. CIBC FCA at para. 37.

21. CIBC FCA at paras. 40-41.

22. CIBC FCA at para. 44.

23. CIBC FCA at para. 45.

24. See Elie S. Roth and Ryan Wolfe, "CIBC v Canada: A Matter of Precedence", International Tax Highlights, Volume 2,, Number 3, August 2023.

25. The Queen v. MacMillan Bloedel Ltd., 99 DTC 5454 (FCA). [MacMillan Bloedel]

26. An exception is subsection 39(3) where a taxpayer that has issued a bond, debenture or similar obligation and repurchases the obligation "in the open market, in the manner in which any such obligation would normally be purchased in the open market by any member of the public". The difference between the issue price and purchase price is deemed to be a capital gain (or capital loss if the purchase price exceeds issue price) from the disposition of a capital property.

27. Ruth Sullivan, Statutory Interpretation, 3^rd. ed. (Toronto: Irwin Law, 2016) at p. 43.

28. Minister of National Revenue v. Consolidated Glass Ltd., [1957] SCR 167. [Consolidated Glass]

29. Consolidated Glass at para. 8-9.

30. Including, for example, in MacMillan Bloedel.

31. Technical Notes to Subsection 39(2), Department of Finance, October 24, 2012.

32. CIBC FCA at para. 31.

To view the original article click here

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.