One limit, allocated on Schedule 23
The mechanics are an annual agreement, not a formula. Each year, the associated group agrees on percentages of the $500,000 business limit for each corporation, files Schedule 23 with each company's T2 showing the same allocation, and each corporation then claims the small business deduction on the least of three numbers: its active business income, its taxable income, and its allocated share of the limit. A company allocated 0% claims nothing regardless of its income; a company allocated 100% can shelter up to the full $500,000 of active business income at the 12.2% combined Ontario rate.
The rules around the edges are strict even though the split itself is free. The percentages must total 100 or less across the group, every corporation must file consistently, and if the group files conflicting schedules or none at all, CRA can impose an allocation of its own choosing, which is never the plan. The limit is also prorated for short taxation years, so a company that just incorporated or changed its year-end gets a fraction of its share, and corporations that become associated partway through their year need the year of change reviewed rather than assumed.
Two pieces of timing fine print round out the mechanics. Corporations share the limit for a year if they were associated at any time in that year, and each company applies the test for its own taxation year, so groups with staggered year-ends need the overlap mapped rather than guessed. And while only a CCPC can claim the small business deduction at all, a member with nothing to claim still counts: its taxable capital and its investment income go into the group totals that grind the limit for everyone else.
Note what is being shared: the limit, not the rate. Both companies remain CCPCs and both file their own returns; association never merges their taxes. What the small business deduction itself does, and which income qualifies, are covered in what the small business deduction is and what counts as active business income; this page is about dividing it.
How to choose the split
Allocate the limit to where the active business income is, and revisit it every year. If OpCo earns all the group's operating profit and HoldCo only holds investments, the split is 100/0 and there is nothing to optimize: a holding company with no active business income has no use for any allocation, because the deduction can never exceed active business income. Most two-company groups in the GTA look like this, and for them Schedule 23 is thirty seconds of decision and one line of consistency to maintain.
The split starts to matter when more than one company earns active income. Two points anchor the decision. First, when the group's combined active income is inside $500,000, any allocation that covers each company's income works, and the group's total tax is the same; the choice mostly decides which company shows the lower tax bill and holds the cash. Second, when combined active income exceeds the limit, the excess is taxed at the 26.5% general Ontario rate somewhere no matter how you split, so the real question becomes which company should carry the higher-taxed income. That answer follows practical facts: where losses might arrive, which company is being groomed for sale or financing, and which province each company pays tax in when the group is not all-Ontario.
A concrete version. OpCo A earns $400,000 of active income, OpCo B earns $250,000: combined $650,000 against one $500,000 limit. One workable split allocates $400,000 of limit to A and $100,000 to B, so B pays the general rate on $150,000 of its profit; another covers B fully and leaves A exposed on the same $150,000. Federally the group's total bill is the same either way; the split decides which company carries the higher-taxed income, which matters when one of them is building retained earnings for a purchase, showing results to a lender, or filing in a different province. Whatever you choose, revisit it annually, because the split that fit last year's income rarely fits this year's.
What the split cannot do is move income. Shifting profit between associated companies takes real transactions, typically management fees or intercompany charges that must be defensible as actual services at reasonable amounts, or a reorganization. Allocating limit toward a company does nothing unless that company genuinely earns the income, which is why limit planning and structure planning travel together.
Two grinds shrink the limit before you split it
The $500,000 the group divides is a ceiling, and two separate reductions can lower it, both measured across the whole associated group. The group applies whichever reduction is larger, then splits what is left.
| Taxable capital grind | Passive income grind | |
|---|---|---|
| Trigger | Combined taxable capital employed in Canada above $10 million | Combined adjusted aggregate investment income above $50,000 |
| Fully gone at | $50 million of combined taxable capital | $150,000 of combined investment income |
| Measured when | Prior year, across all associated corporations | Prior year, across all associated corporations |
| Pace | Straight-line as capital rises through the range | The limit falls $5 for every $1 of investment income over $50,000 |
| Ontario parallel? | Yes, Ontario's limit grinds the same way | No, Ontario did not adopt this grind |
The taxable capital grind catches larger, asset-heavy groups: think substantial equipment, real estate held in the group, and retained earnings across several companies. The passive income grind catches successful groups of any size that have accumulated an investment portfolio. Both are group tests precisely so that owners cannot dodge them by parking capital or portfolios in a separate associated company.
They also behave differently in planning. Taxable capital moves slowly, so a group drifting toward $10 million can see it coming years out and shape structure and debt accordingly. Investment income can jump in a single year, one good capital gain is enough, which makes the prior-year measurement both a warning system and a trap: this year's portfolio decisions set next year's limit.
The passive income grind is a group test, and holdcos are the usual culprit
The most common way an operating company loses its small business rate is a holding company's investment account. Adjusted aggregate investment income captures, broadly, the group's interest, taxable capital gains from passive investments, rents and portfolio dividends, measured in the prior year. Once the group total passes $50,000, every dollar over it removes five dollars of the shared limit, and at $150,000 the federal limit is gone entirely: the operating company pays the general federal rate on all of its profit even though the portfolio sits elsewhere.
What counts is broader than a brokerage statement but narrower than all investment activity. In, broadly: interest, taxable capital gains from passive investments net of losses, rents from passive holdings, and portfolio dividends. Out: gains on assets used in the active business, income that is incidental to the business such as interest on a working-capital float, and dividends from connected corporations in the group, which are handled under their own rules. Where a number sits on that line is worth confirming rather than assuming, since each dollar in costs five dollars of limit.
Three softeners keep this from being as bad as it first sounds. Ontario did not parallel the grind, so a corporation that loses the federal small business rate this way keeps Ontario's 3.2% small business rate on its first $500,000, which blunts the jump. The heavy corporate tax on the investment income itself is partly refundable when dividends are paid, a mechanism we walk through in how RDTOH and the dividend refund work. And because the test looks at the prior year, a one-time spike, say a large capital gain on selling a property, grinds one year's limit and then releases it.
Planning responses exist, but they are owner-level decisions rather than tricks. Paying more salary reduces active income exposed to the grind and builds RRSP room; paying dividends moves capital out of the corporation so it stops generating passive income at all; realizing gains gradually rather than in one year manages the spike; and reviewing what the portfolio holds changes how much of its return counts as investment income. Lending surplus from the operating company to the holding company changes nothing here, because the group is tested together wherever the money sits. Which levers are worth pulling depends on your personal tax picture as much as the corporate one, which is exactly the trade-off a corporate tax planning CPA in Ontario should be modelling with you before year-end, not after.
Filing it right, and the mistakes we see
Getting this wrong is usually a filing failure, not a planning failure. The recurring mistakes in files that come to us:
- Two full limits claimed because each company has its own accountant and neither asked about the other company. Every open year is exposed to reassessment with interest.
- Association never re-tested after a share sale, a marriage, a trust being added, or options granted, so the group composition on Schedule 9 is stale.
- The holdco's portfolio ignored when computing the grind, because the operating company's preparer never sees the investment statements.
- Inconsistent Schedule 23s across the group, or an allocation that no one updated when the income moved.
- Short years and mid-year association changes handled as if nothing happened, missing the proration.
- Ontario left on the table when a federally ground group assumes the provincial rate is gone too and stops claiming the Ontario small business deduction it still qualifies for.
The fix is structural: one advisor who sees the whole group, tests association annually, computes the grinds on combined numbers, and files every company's schedules from a single working paper. That is how we run multi-company groups inside our corporate tax work, with the allocation decided at planning time rather than discovered at filing time.
What changes the answer for your group
Whether sharing the limit costs you anything, and how to split it, comes down to a handful of facts:
- Whether the corporations are actually associated, which is its own analysis of control, family and cross-ownership, covered in what associated corporation status is
- Where the active business income sits, since allocation follows income and a holdco needs none
- Combined profit versus $500,000, because below it the split is low-stakes and above it someone pays the general rate regardless
- The group's prior-year investment income, including every associated company's portfolio, against the $50,000 threshold
- Combined taxable capital against the $10 million threshold for asset-heavy groups
- What each company is being positioned for, since financing, sale or succession plans can outrank a small rate difference
Run those once a year, at planning time rather than filing time, and the limit takes care of itself; settling the allocation before year-end also leaves room for salaries and bonuses to move income between companies legitimately while it still counts. If your companies have been filing as strangers, or a portfolio has grown quietly inside a holdco, the group-level math in corporate tax planning for owner-managed businesses is where we would start, and a free 15-minute discovery call is enough to tell you whether the current filings hold up.
