Use of high-PUC, low ACB shares to create CDA (pp. 4-5)
[T]he new rules denied the creation of ACB in shares issued on a stock dividend or pursuant to a PUC bump unless the taxable dividend was paid out of safe income on hand of the dividend payer. [fn 30: Subclause 52(3)(a)(ii)(A)(II) for a stock dividend, and subparagraph 53(1)(b)(ii) for a PUC bump.] …
The ACB denial rules can cause a parent's shares of its subsidiary to have mismatched tax attributes. For a stock dividend or a PUC bump to which the ACB denial rules apply, the shares of the subsidiary can have higher PUC than ACB. Where the redemption price of the shares is equal to their PUC, the parent may cause such shares of its subsidiary to be redeemed in order to realize a capital gain. Alternatively, the parent might dispose of those shares of the subsidiary in another manner, whether to a third party or pursuant to an internal reorganization, in order to realize the capital gain. In either case, on any disposition of its high-PUC, low-ACB shares of the subsidiary, the parent may increase its CDA by virtue of its realized capital gain. This enables a "mixing-and-matching" distribution of the parent's pre-existing safe income on hand from one source, such as its active business income, to an individual shareholder as a capital dividend, using its CDA created from a capital gain realized on any disposition of its high-PUC, low-ACB shares of the subsidiary.
Purpose of CDA penalty rule in s. 89(1) – CDA – (a)(i)(A) (pp. 5-6)
In recognition that the ACB denial rules created this surplus stripping opportunity,…[t]he CDA penalty denies the creation of CDA in respect of the portion of a realized capital gain that is preserved by the ACB denial rules….
[I]t is possible to argue that the CDA penalty recalculates ACB for purposes of CDA only in order to frustrate more modern surplus stripping. Where the ACB denial rules apply, the subsidiary's safe income on hand is converted into ACB, reducing the parent's built-in capital gain in its shares of the subsidiary. The parent may dispose of such shares of the subsidiary in order to realize a capital gain that was preserved by the ACB denial rules. The tax-free portion of this gain is included in the parent's CDA. As a result, the parent may make a "mixing-and-matching" distribution of its own pre-existing safe income on hand to an individual shareholder as a tax-free capital dividend.
Example illustrating punitive effect of CDA penalty rule (p. 6)
[A]n individual owns all of the shares of a private corporation resident in Canada (“Holdco”). Holdco owns all of the shares of another Canadian-resident corporation ("Opco") having a fair market value of $100, PUC of nil, and ACB of nil. Thus, Holdco's built-in capital gain in its Opco shares is $100. Assume that Opco's safe income on hand attributable to its shares held by Holdco is $25. Imagine that Opco miscalculated its safe income on hand attributable to its shares held by Holdco. Thus, Opco undertakes a PUC bump of $100, giving rise to a deemed dividend of $100.
On the PUC bump, the ACB to Holdco of its Opco shares does not increase by the full amount of the deemed dividend of $100. Rather, Holdco's ACB in the Opco shares increases only to the extent that the PUC bump is paid out of safe income on hand of $25. This more modest increase in ACB reduces Holdco's built-in capital gain in its Opco shares to $75, preserving the portion of the capital gain that was not attributable to safe income on hand….
[S]olely for calculating Holdco’s CDA, the ACB to Holdco of its Opco shares is increased by the full amount of the deemed dividend of $100. Therefore, any subsequent disposition of the Opco shares by Holdco does not give rise to a capital gain for purposes of computing Holdco's CDA. This is a punitive measure; on such a subsequent disposition, the portion of Holdco's realized capital gain that was preserved by the ACB denial rules cannot increase its CDA.
Avoidance of CDA penalty by setting PUC of stock dividend shares below estimated safe income (pp. 18-19)
[C]onsider a high-low stock dividend that is paid on common shares of a dividend payer that are held by the dividend recipient, consisting of preferred shares with a high redemption price (and thus fair market value) of $100, but a low stated capital (and thus PUC) of $40. The dividend payer's safe income on hand attributable to its common shares on which the stock dividend is received is $70….
By setting PUC of the issued shares below estimated safe income on hand, the CDA penalty can be avoided regardless of whether subsection 55(2) recharacterization is respected when determining the dividend "amount". In this example, subsection 248(1) sets the "amount" of the high-low stock dividend at the PUC of the preferred shares of $40. Subsection 55(2.2) adjusts this "amount" of the high-low stock dividend to $100 for the stipulated purposes in section 55. Subsection 55(2.3) bifurcates this $100 adjusted dividend "amount" into a safe income dividend of $70, and non-safe income dividend of $30, solely for purposes of section 55.
The methodology to compute the ACB to the dividend recipient of the preferred shares of the dividend payer issued on the stock dividend turns on whether the dividend "amount" is $40 as computed under subsection 248(1), or only $10 comprising the portion of that same dividend "amount" that remains after recharacterization under subsection 55(2). In each case, the ACB to the dividend recipient of the preferred shares is $100, and the CDA penalty does not apply. …
A high-low stock dividend having a redemption price (and thus fair market value) above safe income on hand, but stated capital (and thus PUC) below safe income on hand, is an optimal way to crystallize safe income on hand. This strategy avoids uncertainty concerning the dividend "amount", and preserves integration by avoiding the CDA penalty. The circumvention of the CDA penalty is not abusive. The reason is that, from a tax policy perspective, the CDA penalty is no longer needed to deter surplus stripping of the dividend payer's safe income on hand after the 2015 amendments.
Introduction of the bifurcation rules (p. 10)
Any other taxable dividend is no longer protected from recharacterization by the internal reorganization exception. Each such taxable dividend does not arise as a foundational component of a sanctioned divisive reorganization but, rather, has the potential, as in D & D Livestock, to effect a significant reduction in a capital gain on any share, a significant reduction in the fair market value of any share, or a significant increase in the total cost of property of the dividend recipient….
However, the holding corporation could forgo a designation under paragraph 55(5)(f). The full amount of the taxable dividend could be recharacterized by subsection 55(2) as a deemed capital gain that would increase its CDA. In that case, a portion of the safe income on hand of the dividend payer attributable to its shares held by the dividend recipient could be realized as a capital gain. This is the very notion of surplus stripping. The holding corporation could then elect to distribute to its individual shareholder another source of income, such as its active business income, as a capital dividend using such CDA generated by the deemed capital gain.
[T]o neutralize that threat, amended paragraph 55(5)(f) (and new subsection 55(2.3) for high-low stock dividends) (collectively, the "bifurcation rules") now compel die distribution of safe income on hand pursuant to a taxable dividend….
Each of the bifurcation rules divides a taxable dividend into a safe income dividend and a non-safe income dividend. As the safe income dividend reduces the dividend recipient's accrued gain in its shares of the dividend payer, the shareholder's CDA cannot be increased to the extent of the amount of that safe income dividend. This ensures that subsection 55(2) cannot transform the portion of a capital gain that is attributable to safe income on hand into a positive balance of CDA for purposes of surplus stripping.
Issue of whether the purpose of a whole taxable dividend is determined before or after bifurcation (pp. 10-11)
If the purpose of the whole taxable dividend under a safe income crystallization must be identified before bifurcation, one might concede that one of the purposes of the dividend…is to achieve that outcome of significantly reducing a capital gain, significantly reducing the fair market value of any share, or significantly increasing the total cost of property of the dividend recipient.
In contrast, if the bifurcation rules divide the whole taxable dividend into a safe-income dividend and a non-safe income dividend before the "purpose" of either divided is ascertained … [o]nly the purpose of the non-safe income dividend … must be ascertained. …
[C]onsider where a taxpayer miscalculates its safe income on hand before undertaking a safe income crystallization. The taxpayer's subjective purpose is not to give rise to a non-safe income dividend pursuant to the bifurcation rules….
[T]he three purpose tests are meant to inquire into the subjective purpose of the dividend payer and the dividend recipient in order to frustrate tax-driven capital gains stripping. Only the first interpretive approach meets that legislative objective….
CRA view consistent with purpose being identified before bifurcation (p. 11)
The CRA's position is that the 2015 amendments did not change the order of application of these rules but, rather, merely made the operation of paragraph 55(5)(f) automatic rather than discretionary. [fn 52: 2017-0726381C6]
[T]he CRA’s view ignores that bifurcation under paragraph 55(5)(f) occurs "for purposes of this section [55]", and that bifurcation under paragraph 55(2.3)(a) occurs "for the purpose of [subsection 55(2)]". This means that automatic bifurcation of the taxable dividend applies for purposes of the three purpose tests. Indeed, this is the very concept that raises the interpretive issue in the first place. Still, despite its unsound reasoning? the CRA came to the proper conclusion on that issue.
Incoherent scheme for divestitures: integration where Holdco sells Opco to arm’s length Holdco; no integration where Opco redeems Holdco shares (pp. 24-25)
To illustrate this incoherent legislative state, consider the following example. "Opco" is owned by three unrelated holding companies, "Holdco A", "Holdco B" and "Holdco C". Each of Holdco A, Holdco B and Holdco C are owned by individuals "A", "B" and "C", respectively. The shares of Opco owned by Holdco A have a fair market value of $100, PUC of nil, and ACB of nil. The safe income on hand attributable to the shares of Opco held by Holdco A is $30. "A" wishes to divest its indirect interest in Opco. Conceptually, a number of options are available, and the tax results differ depending on the structure chosen for the divestiture:
- "A" sells the shares of Holdco A to "B" and realizes a capital gain of $100, resulting in personal tax at the rate applicable to capital gains realized by individuals.
- Holdco A sells the shares of Opco to Holdco B and realizes a capital gain of $100. Although there is safe income of $30, the bifurcation rules do not apply because Opco did not pay a taxable dividend to Holdco A. Thus, Holdco A's capital gain augments its CDA. Holdco A can distribute its sale proceeds to A as a capital dividend and a noneligible dividend to achieve an effective combined tax rate on the distribution that approximates that applicable to capital gains realized directly by an individual.
- Opco redeems its shares held by Holdco A. A triggering event under paragraph 55(3)(a) occurs, since unrelated persons have an increase in interest in Opco. On the share redemption, subsection 84(3) gives rise to a deemed dividend. Paragraph 55(5)(f) bifurcates the resulting deemed dividend into a safe income dividend of $30 and a non-safe income dividend of $70, solely for purposes of section 55. The safe income dividend is excused from recharacterization as a deemed capital gain pursuant to the safe income exception. However, subsection 55(2) recharacterizes the non-safe dividend as a deemed capital gain in the amount of $70. Only this smaller quantum of the non-safe income dividend recharacterized as a deemed capital gain augments Holdco A's CDA. As the safe income dividend reduced the quantum of the deemed capital gain, Opco's safe income on hand that is attributable to its redeemed shares cannot be surplus stripped. The share redemption is treated as a corporate distribution that is taxed differently than a sale of its shares.
Consequently, the statutory regime governing divestitures is not consistent. Share redemptions caught by subsection 55(2) produce the most punitive results….
The Department of Finance (Canada) sought, through its withdrawn proposed section 246.1, to deny CDA created on an internal reorganization that was not caught by subsection 55(2) but which could have facilitated surplus stripping. Upon the abandonment of that measure, there is no clear tax policy against surplus stripping achieved through realizing the economic value of a corporation's safe income on hand by a sale of shares that does not invoke the bifurcation rules. Further, withdrawal of proposed section 246.1 indicates that there is no abuse from surplus stripping through so-called "mixing-and-matching" distribution.
Two main competing interpretations of the “amount” of a stock dividend for purposes of s. 52(3)(a)(ii)(A) (pp. 13-15)
To apply the ACB computation rules for stock dividends, one must ascertain the “amount” of the stock dividend….there are five potential dividend “amounts” to consider.
First, subsection 248(1) defines the “amount” of the stock dividend as the PUC of the issued shares. This “amount” applies generally throughout the Act. …
Fifth, where subsection 55(2) applies to a non-safe income dividend, paragraph 55(2)(a) deems the "amount" not to be a dividend received by the dividend recipient and paragraph 55(2)(c) deems the "amount" to be a capital gain of the dividend recipient from the disposition of a capital property. Subsection 55(2) applies "notwithstanding any other provision of this Act". Therefore, it is possible that another dividend arises for purposes of the entire Act, consisting of the difference between the "amount" of the stock dividend determined under subsection 248(1) and the portion of that taxable dividend that is deemed by paragraph 55(2)(a) not to be a dividend….
The Department of Finance (Canada) considers that the "amount" of the stock dividend, under the ACB computation rules, is the dividend "amount" under subsection 248(1). This interpretation of the "amount" of the stock dividend for purposes of the ACB computation rules does not respect recharacterization under subsection 55(2). Where recharacterization applies, paragraph 55(2)(a) deems the "amount" of the dividend not to be a dividend received, and paragraph ,55(2)(c) deems the "amount" of the dividend to be a capital gain….
If the interpretation of the Department of Finance (Canada) prevails, the ACB denial rule can apply. The reason is that a portion of the subsection 248(1) "amount" of the stock dividend could exceed safe income on hand. As a consequence, the CDA penalty applies. This leads to under-integration….
Under-integration is eliminated where the "amount" of the stock dividend for purposes of the ACB computation rules is interpreted as the fifth possibility. Under this construction, the "amount" of the stock dividend for purposes of the ACB computation rules ought to be the difference, if any, between the subsection 248(1) dividend "amount" and the portion thereof that is recharacterized as a deemed capital gain….
[T]he Department of Finance (Canada) could study whether a simple rule could deem the amount of the safe income dividend to be the relevant "amount" of the stock dividend for purposes of ACB computation (and indeed for all other purposes of the Act). In that way, recharacterization of the non-safe income dividend as a deemed capital gain would be respected throughout the Act, as directed by the text of subsection 55(2).
Issue under s. 52(3)(a)(ii)(A) and (B) - B of the amount of a high-high stock dividend (pp. 16-17)
[I]magine a high-high stock dividend is paid on common shares, comprising the issuance of preferred shares with a high redemption price (and thus fair market value) of $100, and an equivalently high stated capital (and thus PUC) of $100. Assume that the dividend payer's safe income on hand attributable to its common shares held by the dividend recipient is only $70.
Subsection 248(1) sets the "amount" of the high-high stock dividend at the PUC of the preferred shares of $100. The special rules governing high-low stock dividends do not apply to adjust this dividend "amount". [fn 68: Paragraph 55(2.4)(b)] Rather, the general bifurcation rule in paragraph 55(5)(f) divides the high-high stock dividend of $100 into a safe income dividend of $70, and a non-safe income dividend of $30, solely for purposes of section 55. …
[A]s the high-high stock dividend implemented a safe income crystallization, one of the purposes of the stock dividend was to reduce a capital gain or the fair market value of common shares of the dividend payer significantly. Therefore, the non-safe income dividend is recharacterized as a deemed capital gain of $30.
The ACB to the dividend recipient of the preferred shares includes the total of the amounts determined under clauses 52(3)(a)(ii)(A) and (B). A critical issue becomes whether the "amount" of the high-high stock dividend is the subsection 248(1) "amount" of $100, or only the portion of that same dividend "amount" that remains after recharacterization in the amount of $70, for purposes of the ACB computation rules. …
If the dividend "amount" were $100, as suggested in the Technical Notes, the total of the amounts determined under clauses 52(3)(a)(ii)(A) and (B) is $100….
Clause 52(3)(a)(ii)(A) includes in ACB of the issued shares only the portion of the high-high stock dividend that is paid out of safe income on hand in the amount of $70. The reason is that the ACB denial rule in subclause 52(3)(a)(ii)(A)(II) applies to deny inclusion in ACB of the issued shares for the portion of the "amount" of the stock dividend that was not paid out of safe income on hand. Under clause 52(3)(a)(ii)(B), the amount of $30 that is determined under the formula "A + B" is added to ACB of the issued shares. In particular, variable "A" includes the amount of the deemed capital gain arising under subsection 55(2) in the amount of $30. Variable "B" computes to nil because paragraph 55(2.3)(b) did not apply.
If the dividend "amount" were only $70, as obtains if recharacterization of the non-safe income dividend as a deemed capital gain under subsection 55(2) were respected, the ACB of the issued shares remains $100….
Clause 52(3)(a)(ii)(A) still includes ACB of the issued shares the portion of the high-high stock dividend that is paid out of safe income on hand in the amount of $70. In this case, however, the ACB denial rule in subclause 52(3)(a)(ii)(A)(II) does not apply. No portion of the dividend "amount" exceeded safe income on hand.
Under clause 52(3)(a)(ii)(B), the amount of $30 that is determined under the formula "A + B" is added to ACB. Variable "A" includes the amount of the deemed capital gain arising under subsection 55(2) in the amount of $30. Variable "B" computes to nil because paragraph 55(2.3) (b) did not apply. …
If the dividend "amount" were $70, respecting recharacterization under subsection 55(2), the ACB denial rule does not apply. Consequently, the CDA penalty is not applicable. Integration is preserved.
In contrast, if the dividend "amount" were $100, the ACB denial rule applies because the dividend "amount" exceeded safe income on hand. Thus, the CDA penalty applies, leading to under-integration.
On that latter interpretation, solely for purposes of computing the dividend recipient's CDA, the ACB of the preferred shares becomes $130. For purposes of CDA only, ACB of the issued shares is computed by ignoring any reduction in cost basis under subclause 52(3)(a)(ii)(A)(II). On subsequent disposition of its preferred shares by the dividend recipient, a capital loss arises that grinds its CDA balance. In effect, after the preferred shares are sold, the CDA balance excludes the tax-free portion of the deemed capital gain arising under subsection 55(2).
Potential double taxation if safe income crystallized by way of PUC bump (p. 20)
Consequently, for a safe income crystallization, where a PUC bump exceeds safe income on 'hand, the non-safe income dividend can be recharacterized by subsection 55(2) as a deemed capital gain. The critical issue becomes whether double taxation arises on a subsequent disposition of the shares, as the ACB to the dividend recipient of its shares does not increase on account of its deemed capital gain.
Before the 2015 amendments, the CRA stated that double taxation would not be sought. [fn 71: 2011-0415891E5 and 9830665]
Email this Content