The loss that a capital dividend ground away

The situation

A business owner died holding all the shares of his operating holding company through his estate. To fund the estate's debts and equalize a bequest to a non-business child, the estate's executor had the company redeem a substantial block of the shares for cash within the estate's first taxation year, funding part of the redemption out of the company's capital dividend account and part out of ordinary retained earnings.

The problem

A share redemption produces a deemed dividend on the difference between the redemption proceeds and the shares' paid-up capital, and separately a capital loss (or gain) measured against the shares' adjusted cost base. Where the estate is a graduated rate estate, s.164(6) lets it elect to carry a capital loss realized “in the course of administering the graduated rate estate of a taxpayer” back to be treated as the deceased's own loss in his terminal year — but only for a disposition “within the first three taxation years of the estate.” The estate here qualified on both counts.

What the estate had not modelled was s.112(3.2), the stop-loss rule for a loss on a share held by a trust. It provides that the amount of the loss “is deemed to be the amount of the loss determined without reference to this subsection minus the total of” a formula built substantially around dividends received on the share — including, on these facts, the deemed dividend the redemption itself generated. The very transaction that created the loss also funded a dividend that reduced it.

The numbers

The redemption produced a capital loss of roughly $1.4 million on the shares before any grind, against a terminal-year capital gain on the deceased's other assets of about $1.1 million. If the full loss had been available, the estate could have wiped out the terminal gain entirely. After the s.112(3.2) grind against the dividends received on the redeemed shares, the loss available for the s.164(6) carryback fell to roughly $740,000 — enough to offset about two-thirds of the terminal gain rather than all of it.

The rule that decided it

The formula is not a blanket denial. s.112(3.2) carries its own carve-out for exactly this fact pattern: where “the trust is an individual's graduated rate estate, the share was acquired as a consequence of the individual's death and the disposition occurs during the trust's first three taxation years,” the grind is reduced by one-half of the lesser of the loss and the deceased's own capital gain on the share immediately before death. That carve-out is what kept roughly half the loss alive; without it, the grind would have consumed most or all of the loss, because the dividends received on the shares were large relative to the loss itself. Choosing between this route and an alternative — a pipeline transaction that avoids triggering a dividend at all — is the actual decision on a file like this, and the two techniques answer to opposite constraints: the pipeline needs time and no immediate cash need; the loss-carryback route needs a graduated rate estate acting inside three years and accepts a partial, ground-down result in exchange for speed.

The outcome

The estate filed the s.164(6) election for the reduced $740,000 loss, offsetting most but not all of the terminal-year gain and leaving a residual tax liability the estate paid from the redemption proceeds themselves. In hindsight, the executor and the estate's advisors concluded a smaller redemption paired with a pipeline transaction for the balance would likely have preserved more of the relief, but the estate's cash-flow needs at the time made the faster, partial route the practical choice.

What it would have cost otherwise

Modelling the s.112(3.2) grind before choosing the redemption amount would not have avoided the grind, but it would have let the executor size the redemption to balance the estate's cash needs against how much loss the grind would leave usable — potentially preserving tens of thousands of dollars more relief for the same cash raised, simply by sequencing which shares were redeemed and how much of the redemption came from the capital dividend account.

The tell

Any post-mortem plan that treats a share redemption's loss and its dividend as two separate line items is missing the mechanic: s.112(3.2) links them directly, so the size of the dividend the redemption produces determines how much of the loss survives to be carried back. The choice between a pipeline and a loss carryback should be made with the grind modelled first, not assumed away because a graduated rate estate and a three-year window are both available.

Takeaways

• • s.164(6) only reaches a loss realized within a graduated rate estate's first three taxation years — the deadline is a hard constraint on this whole technique.
• • s.112(3.2) grinds that loss by dividends received on the shares, including the deemed dividend the redemption itself produces.
• • A carve-out preserves roughly half the loss, up to the deceased's own terminal gain, specifically for a graduated rate estate acting inside the window.
• • Model the grind before sizing the redemption; a pipeline may preserve more value where time allows.

Sources

• • Income Tax Act — s.164(6), the graduated rate estate loss carryback to the deceased's terminal year.
• • Income Tax Act — s.112(3.2), the stop-loss rule reducing a trust's loss on a share by dividends received, with the graduated rate estate carve-out in subparagraph (iii).
• • Treadstone Law — how the stop-loss rules affect a capital loss — general background on stop-loss mechanics. It does not address the graduated-rate-estate carve-out that determined the outcome here.
• • Not verified: the exact dollar figures above are illustrative of this composite and were not derived from a published CRA worked example; the mechanism — the grind and the carve-out — is what the statutory citations support.