Anonymised, illustrative composite. A retiring founder sold his refrigeration and HVAC service company; the money stayed behind in a holding corporation, and the only reason his spouse's dividend from it escaped the top-rate split-income tax was his own age.
At a glance
A single owner built a two-location commercial refrigeration and HVAC service company over twenty-six years. He sold the operating business in an asset sale for $2,850,000. The vendor corporation did not wind up — it kept the after-tax proceeds, retired its old service contracts, and became a plain investment holding company, buying a laddered portfolio of provincial and corporate bonds. Its shareholders did not change: the founder held 84% of the common shares, his spouse held 6% (subscribed for with her own funds when the company was incorporated, not a gift or a loan from him), and an adult son who had never worked in the business held the remaining 10%.
Since 2018, ITA s.120.4 has added a top-rate tax on “split income” received by a specified individual — broadly, a family member of a business owner — whether or not that family member had anything to do with earning it. A taxable dividend on shares of a private corporation falls inside the definition directly: split income includes an amount required to be included in computing the individual's income in respect of “taxable dividends received by the individual in respect of shares of the capital stock of a corporation”, other than shares listed on a designated stock exchange. Holdco's shares are not listed anywhere.
The year after closing, Holdco's bond ladder had matured enough interest that the board declared a dividend, and the spouse's 6% block produced $58,000 of it. The firm's first read was that it would be taxed at the “highest individual percentage” under s.120.4(2), because she did not independently clear the exclusions built into the split-income rules. “Excluded shares” requires the specified individual to hold, immediately before the dividend, shares that give at least 10% of the votes and at least 10% of the fair market value of the corporation; at 6% she failed both. A “reasonable return” argument was not realistic either — she had never worked in the business, had contributed no property beyond a decades-old share subscription, and had assumed no risk since. Whatever amount that test might defend was unlikely to reach $58,000.
$58,000 of dividend income, against a purchase price of $2,850,000 and a Holdco portfolio of roughly $2.1M after the deal's closing costs and taxes on the sale itself. Nothing about the dollar amounts decided the outcome. What decided it was a single fact that had nothing to do with the balance sheet: the founder's date of birth.
s.120.4(1.1) is titled “Additional rules — specified individual,” and paragraph (c) does the work. An amount is deemed to be an excluded amount of the specified individual for the year if “the following conditions are met: (A) the amount would be an excluded amount in respect of the specified individual's spouse or common-law partner for the year, if the amount were included in computing the spouse or common-law partner's income for the year, and (B) the spouse or common-law partner has attained the age of 64 years before the year”.
Read the unit. The test is not about the recipient at all. It asks a hypothetical: if the founder himself had received this dividend, would it have been excluded for him? On his own 84% block, the answer is yes. “Excluded shares” requires that “less than 90% of the business income of the corporation” came from services — Holdco's is bond interest, so that is trivially satisfied — that the corporation is not a professional corporation, and that the founder clears the 10%-votes/10%-value threshold, which his 84% does many times over. On those facts, a dividend on his shares would be “income from, or a taxable capital gain from the disposition of, excluded shares of the individual” — an excluded amount under paragraph (g)(i) of the “excluded amount” definition. Once that hypothetical clears, condition (B) is arithmetic: the founder had turned 65 five months before the taxation year began, comfortably past “attained the age of 64 years before the year.” Both conditions met, the same $58,000 — though it was actually paid to the spouse, on her own 6% block, which independently qualifies for nothing — is deemed an excluded amount for her.
The shares themselves were never an attribution problem, and it is worth saying why not: she had subscribed for them with her own money at incorporation, so the income-attribution rule in ITA s.74.1 — which reaches shares a spouse acquired by gift or by a loan on which no market interest is charged — was never engaged. TOSI and attribution are separate regimes, tested on separate facts, and this file is only about the first.
The accountant filed the year's return treating the $58,000 as an excluded amount in the spouse's hands, on the strength of s.120.4(1.1)(c). It was taxed at her own marginal rate rather than at the highest individual percentage — a structural difference, not a percentage this file states, since the exact gap depends on brackets this file did not verify for the year in question. Nothing else about the deal or the shareholdings needed to change.
The test is a cliff, not a phase-in. Had the dividend been declared eleven months earlier, before the founder's birthday, s.120.4(1.1)(c)(i)(B) would not have been met on that year's facts and the same $58,000 would have gone to the spouse at the highest individual percentage in full, regardless of how modest her own income was otherwise. The only fix available at that point would have been to wait — there is no partial relief for a source individual who is 64 and eleven months old.
The firm's standard playbook for a case like this reaches for restructuring first — reallocate votes and value so the recipient's own shares clear the 10% excluded-shares threshold, or build a paper trail for a reasonable-return claim. Here the cheapest and safest answer was checking a date of birth before doing either. Before restructuring a family shareholding to solve a TOSI problem, confirm whether the source individual's age already solves it.
Canadian small-business transaction data is not published anywhere, so most valuations in this country quote an American benchmark. The Deavo–Treadstone Acquisition Index is a daily record of Canadian listings built to replace that: asking-price distributions by province and city are published now, and days on market, departure rates and asking-to-sale spreads follow as the series lengthens. Leave an email and we will tell you as each measure lands.
No pitch, no listings. One email when the first report lands.
