Section 55, Safe Income, and Intercorporate Dividends — Law, Lore, and Audit Reality

The dividend was paid.
The tax return was filed.
The deal closed.
Everyone moved on.

Five years later, a CRA reassessment arrived—and with it, a tax bill no one saw coming.

This is how section 55 usually enters the story of a Canadian private business: not at the planning stage, not at filing time, but long after the transaction is finished, when the ability to fix the problem is gone. What looked like a routine intercorporate dividend is suddenly recharacterized as a capital gain. The integration assumptions collapse. And what was once “safe” becomes the most expensive line item in the file.

Section 55 of the Income Tax Act is not obscure. It is not new. But it remains one of the most misunderstood—and professionally dangerous—provisions affecting family-owned enterprises, corporate reorganizations, pipelines, estate freezes, and pre-sale planning in Canada. The risk is not that advisors do not know section 55 exists. The risk is that many still treat safe income as an accounting exercise, rather than what it has become in law and in audit reality: a legal, economic, and evidentiary question.

This article is written to close that gap.

It explains, in clear terms, what section 55 actually does, how CRA now audits safe income, and why historical spreadsheets, retained earnings proxies, and administrative folklore are no longer enough. It is designed to be readable for business owners and non-tax professionals, while remaining technically rigorous enough for tax lawyers, senior CPAs, and advisors who must defend their work years later under scrutiny.

What This Blog Covers

This is a flagship, research-grade resource that moves deliberately from context to statute, from case law to CRA audit behaviour, and finally to execution. Specifically, it includes:

• A Case Brief & Executive Summary
  A concise, skimmable overview for lawyers and senior professionals that frames the legal issue, CRA’s current position, and the planning risk at stake—designed to establish credibility immediately and orient the reader before diving into detail.
• Why Section 55 Still Matters in 2026
  An explanation of why section 55 problems surface late, why reassessments are often irreversible, and why CRA scrutiny has intensified—especially for private and family-owned enterprises.
• What the Act Actually Says (and What It Does Not)
  A statutory reset that separates law from folklore, including the real role of subsections 112(1), 84(3), Part IV tax, and the central importance of paragraph 55(2.1)(c).
• Safe Income vs. Safe Income on Hand
  A clear explanation of what safe income is, what it is not, why adjustments are strictly limited, and how “on hand” should be understood today in light of the courts and CRA audit practice.
• The Robertson Rules, Case Law, and the Kruco Inflection Point
  How historical heuristics emerged, why courts tolerate starting points but not conclusions, and why Kruco remains binding in 2026.
• CRA’s Current Audit Lens (2025–2026)
  How CRA now approaches section 55 audits, what it focuses on in practice, where disputes arise, and where CRA tends to overreach.
• A CRA-Defensible Safe Income Calculation Template
  A step-by-step methodology, with sample tables and documentation guidance, designed to survive audit, objection, or litigation—not just filing.
• Real-World Transaction Context
  Where section 55 issues actually arise: internal family buyouts, pipelines, estate freezes, and pre-sale surplus extraction—and why problems surface after closing.
• Practical Takeaways for Lawyers and Advisors
  When section 55 must be addressed explicitly, when independent review is required, and when historical practice becomes professional risk.

The objective of this article is not to make section 55 sound intimidating. It is to make it understandable, predictable, and defensible—before the reassessment arrives.

 

 

1. Why Section 55 Still Matters in 2026
   (Context, urgency, and audience framing)

Section 55 of the Income Tax Act is one of those provisions that sophisticated private-company owners and experienced advisors “know about,” but too often treat as background noise until it becomes the only thing that matters. In 2026, that risk profile has not softened. If anything, it has sharpened, because CRA’s practical audit posture has evolved toward a more direct, statutory-first application of subsection 55(2) and paragraph 55(2.1)(c), and because the transactions where section 55 bites hardest—internal reorganizations, pipelines, surplus extraction, and pre-sale cleanups—remain routine in family-owned enterprise planning.

This is not a provision that typically fails loudly at the time a return is filed. It fails quietly, later, and expensively—often long after the deal team has moved on, corporate records have gone cold, and the taxpayer’s ability to “fix” the plan has evaporated.

1. Section 55 as a “quietly catastrophic” provision

Section 55 is quietly catastrophic because it usually does not announce itself when the plan is implemented. The dividend gets paid. The intercorporate dividend deduction gets claimed under subsection 112(1). The corporate group continues operating. There may be no immediate Part I tax consequence, and there is rarely a clean line item on the return that screams “high risk.”

The catastrophe arrives later, when CRA reframes what happened: the dividend is treated not as a dividend, but as a gain from the disposition of capital property (or proceeds of disposition in a redemption context), because subsection 55(2) recharacterizes the dividend amount in the circumstances described in subsection 55(2.1).

Why does it rarely surface at filing time? First, safe income is not a T2 schedule you can “tick and tie” without judgment. The core statutory question is whether the dividend exceeds the income earned or realized that could reasonably be considered to contribute to the capital gain inherent in the shares immediately before the dividend. That is a valuation-linked, fact-driven inquiry. The return filing process is not designed to force that analysis, and many corporate groups rely on legacy spreadsheets or accounting proxies that are not built to answer a “reasonable contribution” test.

Second, section 55 is often triggered by series planning: reorganizations, internal steps, and “standard” dividend moves intended to tidy up paid-up capital, isolate assets, or facilitate a future sale. The “safe-income determination time” concept can cut off the safe income measurement period earlier than many expect, which can turn an otherwise defensible safe income narrative into a statutory trap.

Third, CRA’s review cycle is not synchronized with your deal cycle. In many files, reassessments occur years later because the audit is initiated when CRA reviews related-party transactions, reorganizations, or post-sale activity—or when it identifies a pattern of intercorporate dividends without supporting documentation. KPMG’s recent “get audit-ready” commentary reflects that CRA is actively requesting safe income calculations, paid-up capital schedules, adjusted cost base support, and the purpose of dividends paid as part of audit activity.

Why are remedies often unavailable once triggered? Because section 55 failures are not “compliance fixes.” They are structural. Once a dividend has been paid as part of a series, you generally cannot unring the bell. Even if you can amend corporate steps, you may not be able to amend the statutory context that caused subsection 55(2) to apply (including the series analysis and the safe-income determination time). And even where an alternative approach could have been implemented, it often requires contemporaneous documentation, valuations, and sequencing that cannot be recreated retroactively in a credible way.

In other words: when section 55 becomes an issue, it is often already too late to make it not an issue. The file becomes about mitigation, defensibility, and evidence.

1. Why CRA scrutiny has intensified

There are three reasons CRA scrutiny is stronger in 2026 than it was in the era when many practitioners first learned safe income.

The first is legislative redesign. The post-2015 framework emphasized a more direct linkage between dividends and capital gains outcomes, including the statutory language in paragraph 55(2.1)(c) that focuses on income that “could reasonably be considered to contribute to the capital gain.” This matters because it gives CRA a cleaner statutory narrative: the safe income exception is not a general fairness concept or an accounting reconciliation; it is a narrowly framed statutory limitation grounded in contribution to inherent gain.

The second is audit selection. Private-company groups are a natural habitat for intercorporate dividends, internal reorganizations, and surplus extraction—exactly the fact patterns where subsection 55(2) becomes relevant. CRA’s audit requests increasingly ask for the underlying mechanics: safe income computations, adjusted cost base schedules, paid-up capital continuity, and the legal purpose of each dividend and step. The implication for planners is straightforward: you should assume your “section 55 safe income” file could be read by an auditor years later, and that the auditor will expect a defensible computation and narrative.

The third is CRA’s renewed focus on surplus extraction strategies, including pipeline-style planning and internal reorganizations that reduce inherent capital gain. Whether a particular pipeline is defensible is a separate question; the point is that CRA’s attention is concentrated where value is being moved, crystallized, or recharacterized inside corporate groups. Section 55 is one of CRA’s most effective tools in that environment because it converts a dividend (which might otherwise be deductible intercorporately) into a capital gain when the statutory tests are met.

1. The disconnect in practice

The practical disconnect is this: many taxpayers and advisors still treat safe income as if it were a historical, accountant-driven surplus concept—something you approximate using retained earnings, tax pools, or a “safe income on hand” schedule updated annually. That approach can produce a number, but it may not produce a defensible number.

CRA’s modern posture is more pointed. The statutory language requires a link between income and inherent gain. That means “retained earnings” is not an answer; it is, at best, a starting point that requires reconciliation to tax income, statutory adjustments, and an explanation of how the income relied upon contributes to capital gain at the safe-income determination time.

This disconnect shows up in three recurring ways.

First, proxies replace computations. Retained earnings is often treated as equivalent to safe income. But retained earnings is an accounting construct, affected by non-cash accruals, fair value adjustments, and financial statement presentation choices. Safe income is anchored to Division B tax income, modified only as permitted under subsection 55(5), and constrained by the “reasonable contribution” test. Those are different languages. In a section 55(2) audit, CRA will not accept “retained earnings” as a substitute for a statutory computation.

Second, historical spreadsheets are treated as evidence. Many corporate groups maintain a safe income workbook that rolls forward annually. If the workbook does not clearly tie to filed tax returns, the safe-income determination time, the relevant shares, and the accrued capital gain, it becomes fragile. CRA’s information requests increasingly seek not only the computation, but also supporting schedules for ACB, PUC, dividend purpose, and step sequencing.

Third, the role of “safe income on hand” is misunderstood. Some practitioners still approach the analysis as a two-stage inquiry where “safe income on hand” is an overarching legal test. CRA has publicly indicated that it is moving away from applying “safe income on hand” as a formal concept, preferring to focus on the statutory “reasonable contribution” analysis tied to the inherent gain. Whether one agrees with CRA’s framing is not the immediate point. The practical point is that, in audit, CRA will press the taxpayer to justify the link between the income relied upon and the gain that exists immediately before the dividend. If your planning file cannot answer that question with both numbers and narrative, you have a defensibility problem.

In short: what many advisors think is a safe income analysis is often a surplus analysis. CRA is auditing a contribution-to-gain analysis.

1. Intended audience

This flagship piece is written for three overlapping groups, each of whom encounters section 55 safe income risk from a different angle.

First, family-owned enterprises. For owner-managed groups, section 55 problems are uniquely damaging because they often arise during generational transitions, internal reorganizations, or pre-sale planning—moments when the family is focused on succession, liquidity, or preserving a legacy. The harm is not just tax cost; it is irreversibility. A section 55(2) reassessment can unwind the economics of a deal long after the family believed the plan was complete.

Second, corporate and commercial lawyers without tax groups. In many files, legal counsel is leading an acquisition, reorganization, or shareholder transition where intercorporate dividends are part of the mechanics. The legal documentation may be impeccable, but safe income is not a corporate law question; it is a tax computation and evidence question. This is precisely the kind of technical risk that a non-tax legal team should be able to identify early and refer out with confidence. A lawyer does not need to compute safe income to recognize when it is material; they need a clear framework to spot when section 55(2) is in play and when a specialized safe income computation is required.

Third, senior accountants and tax managers. For practitioners, section 55 safe income is not an academic topic; it is a file-quality test. It demands comfort with statutory interpretation, computation discipline, and audit defense. It also demands judgment on issues like safe-income determination time, series analysis, and the evidentiary standard CRA will apply years later. The goal is not only to compute safe income, but to prepare an analysis that can survive scrutiny without relying on folklore, proxies, or undocumented assumptions.

In the sections that follow, we move from context into statute, then into doctrine, then into CRA audit reality, and finally into a CRA-defensible safe income calculation template. The objective is not simply to “explain section 55.” It is to provide a research-grade resource that professionals can use to plan, review, and defend intercorporate dividend transactions where subsection 55(2) is a live risk.

 

1. The Statutory Framework: What the Act Actually Says

(Resetting analysis away from folklore and shorthand)

Any meaningful discussion of section 55 and safe income must begin—and repeatedly return—to the statutory text. Much of the confusion, misapplication, and audit exposure surrounding section 55 arises not from ambiguity in the Act, but from decades of administrative shorthand, accounting proxies, and planning conventions that have drifted away from the words Parliament actually enacted.

Section 55 is not an accounting provision. It is not a surplus management rule. It is not a general anti-avoidance rule in disguise. It is a targeted recharacterization mechanism that operates within a carefully constructed statutory framework governing intercorporate dividends, deemed dividends, and refundable taxes. Understanding that framework is the foundation for any defensible safe income analysis.

1. Intercorporate dividends in context

Intercorporate dividends occupy a privileged place in Canadian tax policy. They are generally deductible, often refundable through Part IV mechanisms, and routinely used in private-company planning. That privilege, however, is conditional. Section 55 exists precisely because Parliament recognized that unrestricted intercorporate dividends could be used to extract corporate value in a way that undermines the capital gains regime.

1. Subsection 112(1): policy objective and limits

Subsection 112(1) permits a corporation to deduct taxable dividends received from a taxable Canadian corporation in computing its taxable income. The policy rationale is integration: income earned in a corporation should not be taxed again simply because it moves up a corporate chain. In the absence of further rules, subsection 112(1) would allow corporate groups to move earnings freely without additional Part I tax.

This policy objective is frequently overstated in practice. Subsection 112(1) is not a blanket endorsement of all intercorporate dividends. It is a mechanical deduction provision that operates subject to other sections of the Act, including section 55. The Act does not say that all intercorporate dividends are acceptable; it says they are deductible unless another provision says otherwise.

This distinction matters because many taxpayers and advisors implicitly equate “deductible” with “safe.” That equation is false. A dividend can be fully deductible under subsection 112(1) and still be recharacterized under subsection 55(2). The deduction and the recharacterization operate at different analytical stages.

A proper section 55 analysis therefore does not ask, “Is the dividend deductible?” It asks, “Is the dividend the kind of dividend Parliament intended subsection 112(1) to protect?” Section 55 is the statutory answer to that second question.

2. Subsection 84(3): deemed dividends on redemptions

Subsection 84(3) deems a dividend to arise when a corporation redeems, acquires, or cancels its own shares for consideration in excess of the paid-up capital of those shares. In private-company planning, redemptions are often used interchangeably with dividends, particularly in buyouts, pipelines, and post-freeze cleanups.

A persistent misconception is that redemptions somehow sit outside section 55 because they are “capital transactions.” Statutorily, the opposite is true. Parliament explicitly deemed the excess redemption proceeds to be a dividend, and that deemed dividend is subject to the same section 55 analysis as an actual dividend.

This is not a technical nuance. Many section 55 failures arise because planners assume that a redemption funded by corporate surplus is less exposed than a cash dividend. In reality, subsection 84(3) is one of the most common gateways through which section 55 applies. If anything, redemptions heighten scrutiny because they often occur in the context of series transactions where the safe-income determination time and the inherent capital gain are actively changing.

In short, redemptions are not a workaround to section 55. They are one of its most frequent entry points.

3. Part IV tax: why refundable tax is not a shield

Part IV tax is often misunderstood as a kind of safety valve. Because Part IV tax on dividends received by private corporations is refundable when dividends are paid onward, some taxpayers assume that paying Part IV tax somehow neutralizes section 55 concerns.

This assumption is incorrect. Part IV tax is an integration mechanism. It addresses timing and rate concerns associated with investment income and dividends flowing through private corporations. It does not validate the character of a dividend for section 55 purposes.

Section 55 asks a different question: whether the dividend is being used, directly or indirectly, to effect a reduction in the inherent capital gain of shares in a manner Parliament did not intend. The presence or absence of Part IV tax is irrelevant to that inquiry. A dividend can attract Part IV tax and still be recharacterized as a capital gain under subsection 55(2).

Treating refundable tax as a shield against section 55 is therefore a category error. The two regimes operate independently, and section 55 overrides the dividend characterization regardless of refundability.

 

1. Section 55 — structure and mechanics

With the intercorporate dividend context in mind, section 55 can be understood for what it is: a targeted rule that overrides the dividend treatment when certain conditions are met, and replaces it with capital gains treatment.

1. Subsection 55(2): the recharacterization rule

Subsection 55(2) is the operative provision. When it applies, all or part of a dividend is deemed not to be a dividend and instead to be proceeds of disposition or a capital gain, depending on the factual context. The provision does not deny the deduction under subsection 112(1); it changes the nature of the amount itself.

This distinction is critical. Section 55 does not “penalize” dividends. It reclassifies them. Once reclassified, the amount is taxed under the capital gains regime, and the downstream consequences—loss of the dividend deduction, potential Part I tax, and disruption of planning assumptions—flow naturally.

2. Purpose test vs results test (post-2015)

Prior to the 2015 amendments, section 55 relied heavily on a purpose-based inquiry: whether one of the purposes of the dividend was to effect a significant reduction in capital gain. That framing invited subjective debate and planning around stated intentions.

The post-2015 framework shifted decisively toward a results-oriented analysis. While purpose remains relevant in certain contexts, the modern structure focuses on whether the dividend has the effect described in subsection 55(2.1). This redesign narrowed the safe income exception and elevated the importance of the contribution analysis in paragraph 55(2.1)(c).

The practical implication is that taxpayers can no longer rely on benign intentions or commercial motivations to insulate a dividend. The statutory inquiry is increasingly mechanical: does the dividend exceed the income that can reasonably be considered to contribute to the inherent capital gain? If yes, subsection 55(2) applies, regardless of why the dividend was paid.

3. The centrality of paragraph 55(2.1)(c)

Paragraph 55(2.1)(c) is the fulcrum of modern section 55 analysis. It frames the safe income exception in precise terms:

income earned or realized by the corporation that could reasonably be considered to contribute to the capital gain on the shares immediately before the dividend.

Every word in this phrase matters.

“Income earned or realized” anchors the analysis in tax income, not accounting income. “Could reasonably be considered” imports an objective, fact-based standard rather than a mechanical formula. “Contribute to the capital gain” ties the inquiry to valuation reality, not surplus accumulation. And “immediately before the dividend” fixes the relevant time, cutting off later developments and retroactive rationalizations.

This is why modern safe income disputes are not won by spreadsheets alone. They are won—or lost—on whether the taxpayer can demonstrate, with evidence and logic, that the income relied upon actually explains the inherent gain in the shares at the relevant time.

 

1. Core statutory reframing

When the statutory framework is read as a whole, a clear reframing emerges—one that runs counter to much of the folklore surrounding safe income.

Section 55 is not about surplus. It does not ask how much accounting profit has accumulated. It is not concerned with whether funds are sitting in a bank account.

Section 55 is not about cash. A corporation can have substantial cash and little safe income, or substantial safe income and little cash. Liquidity is not the statutory test.

Section 55 is not about accounting equity. Financial statement presentation, retained earnings, and book surplus may be informative, but they are not determinative.

Section 55 is about attribution of capital gain. Specifically, it is about identifying how much of the inherent gain in a share is supported by income that Parliament is prepared to let move through the intercorporate dividend system without conversion into a capital gain.

This reframing is not merely academic. It explains why CRA audits have become more focused on narrative coherence, valuation linkage, and contemporaneous documentation. It also explains why many historical planning approaches fail under modern scrutiny. If a safe income computation cannot answer the question “how did this income create this gain?”, it is vulnerable—no matter how neatly it reconciles to retained earnings.

In the sections that follow, this statutory framing will be carried forward into the case law, CRA’s current audit posture, and ultimately into a calculation template designed to withstand scrutiny not just at filing time, but years later, when section 55 tends to surface.

 

III. Safe Income vs. Safe Income on Hand

(Legal meaning versus administrative shorthand)

Few concepts in Canadian corporate tax have generated as much confusion—and as many failed assumptions—as the distinction between safe income and safe income on hand. The two phrases are often used interchangeably in practice, sometimes even within the same file. Yet they are not synonyms, they do not arise from the same source, and they do not perform the same function in a section 55 analysis.

This section resets the analysis to first principles. It explains what safe income actually is, why the Act strictly limits adjustments to it, what “safe income on hand” is not, and how the concept of “on hand” should be understood today—particularly in light of Kruco and CRA’s modern audit posture. The goal is not to rehabilitate administrative shorthand, but to place it in its proper, subordinate role beneath the statutory contribution test.

 

1. Safe income — what it actually is

At its core, safe income is a statutory construct, not an accounting one. It exists solely because section 55 requires a mechanism to distinguish dividends that merely distribute income already reflected in share value from dividends that strip out inherent capital gain.

Safe income begins with Division B, Part I income of a corporation. This is a deliberate legislative choice. Parliament did not anchor safe income in retained earnings, book income, or cash flow. It anchored it in taxable income computed under the Act, subject to specific and limited statutory modifications.

This starting point has three important implications.

First, safe income is corporation-specific. It is computed at the level of the corporation that earned the income. In a corporate group, income does not magically pool across entities. Each corporation’s safe income must be computed independently, and the availability of that income to support a dividend depends on the structure of share ownership and the application of section 55 at each level.

Second, safe income is share-specific. Section 55 asks whether income contributes to the capital gain on particular shares. That means safe income is not a global corporate attribute; it must be linked to the shares in respect of which the dividend is paid. Different classes of shares—particularly in post-freeze or reorganization contexts—may reflect different economic entitlements and different inherent gains. A computation that does not respect share-specific economics is unlikely to withstand scrutiny.

Third, safe income is time-bounded. The statute fixes the inquiry at the safe-income determination time, which is generally immediately before the dividend is paid. Income earned after that moment is irrelevant. Income earned before that moment may or may not be relevant, depending on whether it still contributes to the inherent capital gain at that time. This temporal boundary is one of the most common sources of error in practice, particularly where multi-step transactions or series planning is involved.

Taken together, these features mean that safe income is not a rolling balance that can be casually updated from year to year. It is a snapshot, taken at a precise moment, of income earned or realized that still explains why the shares are worth more than their adjusted cost base.

 

1. Adjustments permitted — and only those permitted

If safe income begins with Division B income, the next question is how that income may be adjusted. The answer is straightforward in principle, even if uncomfortable in practice: only as expressly provided by statute.

Subsection 55(5) sets out the permitted adjustments to safe income. It is not illustrative. It is not advisory. It is exhaustive. Courts have repeatedly emphasized that if Parliament intended additional adjustments to be made, it would have said so. The absence of an adjustment is not a gap to be filled by administrative policy or professional intuition.

This point cannot be overstated. Many historical safe income computations rely on implied adjustments—items added back or deducted because they “feel right” from an accounting or fairness perspective. Examples include adjustments for unclaimed capital cost allowance, notional accruals, or reserves that affect book income but not taxable income. These adjustments may produce a number that aligns with economic intuition, but they do not align with the statute.

The courts’ rejection of implied adjustments reflects a deeper principle: safe income is not designed to measure economic profit in the abstract. It is designed to measure the portion of inherent capital gain that Parliament is prepared to allow to be extracted as a dividend without triggering capital gains treatment. That is a legal policy choice, not an accounting exercise.

This is why judicial decisions culminating in Kruco rejected attempts to retrofit administrative rules into the statute. Adjustments must be grounded in the text of subsection 55(5) or they do not belong in the computation. The fact that an adjustment would produce a more “reasonable” outcome in a particular case is not a sufficient basis to include it.

For practitioners, the practical implication is clear: if you cannot point to a specific statutory provision authorizing an adjustment, that adjustment is vulnerable. In a CRA audit, the burden will be on the taxpayer to justify every deviation from Division B income. Unsupported adjustments are not merely aggressive; they are indefensible.

 

1. Safe income on hand — what it is not

The phrase “safe income on hand” does not appear in the Act. It emerged as administrative shorthand, intended to capture the intuitive idea that income must still exist in some form to support share value. Over time, however, the shorthand hardened into a quasi-rule, often misunderstood and misapplied.

To understand what safe income on hand is, it is first necessary to understand what it is not.

Safe income on hand is not retained earnings. Retained earnings is an accounting residual, shaped by financial reporting standards, management estimates, and presentation choices. It may include non-cash accruals, unrealized gains, and reserves that have no immediate relationship to taxable income or capital gain. While retained earnings may provide context, they cannot substitute for a statutory safe income analysis.

Safe income on hand is not cash. A corporation may have substantial cash on its balance sheet for reasons entirely unrelated to income—borrowings, asset dispositions, or capital contributions, for example. Conversely, a corporation may have little cash because income has been reinvested in non-liquid assets that nonetheless support share value. Cash is a liquidity concept; safe income is a gain-attribution concept.

Safe income on hand is not accounting surplus. Like retained earnings, accounting surplus reflects financial reporting conventions, not statutory intent. The Act does not ask whether surplus exists; it asks whether income contributes to capital gain. Those inquiries may overlap, but they are not coextensive.

The danger of treating safe income on hand as one of these proxies is that it shifts the analysis away from the statutory question. When a practitioner says, “there is enough retained earnings,” or “the company has plenty of cash,” they may be answering an accounting or liquidity question—but not the section 55 question.

 

1. What “on hand” actually means

If safe income on hand is not retained earnings, cash, or accounting surplus, what does it mean?

Properly understood, “on hand” refers to economic survivability. It asks whether the income relied upon in the safe income computation still exists, in some form, to support the value of the shares at the safe-income determination time.

This is not a separate legal test. It is an economic filter applied to the statutory safe income amount. The logic is intuitive: income that has been completely dissipated—through taxes, dividends, losses, or value-eroding expenditures—cannot reasonably be said to contribute to the capital gain inherent in the shares.

The concept therefore focuses on capacity to support share value, not on physical form. Income may be “on hand” even if it has been reinvested in assets, used to reduce debt, or deployed in the business. Conversely, income may be “off hand” even if it once appeared in taxable income, if it has since been consumed in a way that permanently eroded value.

This is where the relationship between safe income and accrued capital gain becomes central. The statutory question in paragraph 55(2.1)(c) is whether the income can reasonably be considered to contribute to the capital gain on the shares immediately before the dividend. The accrued capital gain provides an external constraint: income that does not explain the gain cannot support the dividend.

Seen this way, “on hand” is not about tracing dollars. It is about explaining value. The inquiry is not, “where did the money go?” but rather, “does the income relied upon still explain why the shares are worth more than their cost base?”

This reframing aligns with both the statutory language and modern audit practice. CRA auditors increasingly ask for explanations linking income to value, rather than mechanical reconciliations to cash or surplus accounts. A computation that demonstrates this linkage—supported by facts and documentation—is far more likely to survive scrutiny than one that relies on shorthand proxies.

 

1. Reconciling Kruco, CRA shorthand, and the statutory contribution test

The final step in clarifying safe income versus safe income on hand is reconciling three sources of authority that are often perceived as being in tension: the language used in Kruco, CRA’s historical administrative shorthand, and the statutory contribution test in paragraph 55(2.1)(c).

Kruco reaffirmed that safe income is rooted in taxable income and that adjustments must be statutorily grounded. At the same time, the decision acknowledged that income must still exist in some form to contribute to share value. This is not an endorsement of administrative overreach; it is a recognition of economic reality.

CRA’s historical shorthand attempted to operationalize this reality through the concept of “safe income on hand.” Over time, however, that shorthand was sometimes applied as if it were a separate legal requirement, leading to over-adjustment and the importation of non-statutory concepts.

The statutory contribution test provides the unifying principle. It asks whether income earned or realized could reasonably be considered to contribute to the capital gain at the relevant time. That question inherently incorporates economic survivability without requiring a separate doctrine.

When these elements are properly aligned, the analysis becomes coherent. Safe income is computed using Division B income, adjusted only as permitted by subsection 55(5). An economic filter is then applied to ensure that the income relied upon still explains the inherent capital gain. Administrative shorthand may assist in organizing that analysis, but it does not replace the statutory test.

For practitioners, this reconciliation has practical consequences. It shifts the focus away from arguing about labels—“safe income” versus “safe income on hand”—and toward building a defensible narrative that connects income, value, and gain. In a section 55 audit, that narrative is often the difference between a resolved file and a recharacterized dividend.

In the next sections, this framework will be carried forward into the discussion of the Robertson Rules, the evolution of case law, and CRA’s current audit approach, before culminating in a calculation template designed to operationalize these principles in a way that can withstand scrutiny years after the dividend is paid.

 

1. The Robertson Rules — Necessary Heuristic or Enduring Distraction?

(Positioned carefully, not dismissively)

Any serious discussion of safe income inevitably encounters what practitioners commonly refer to as the Robertson Rules. For decades, these rules have shaped how accountants and tax advisors think about safe income, how spreadsheets are built, and how internal reviews are conducted. They have also contributed, sometimes quietly and sometimes dramatically, to the very misunderstandings that section 55 was designed to prevent.

To dismiss the Robertson Rules outright would be ahistorical and unhelpful. To treat them as binding law is equally dangerous. The correct approach in 2026 lies between those extremes: understanding why the rules emerged, why they remain useful in a limited sense, and why courts and CRA will not accept them as substitutes for statutory analysis.

 

1. Origins of the Robertson Rules

The Robertson Rules did not emerge in a vacuum. They were a response to a genuine policy problem that existed long before section 55 took its modern form.

The pre-1981 surplus stripping problem

Before the comprehensive surplus-stripping regime was introduced, corporate taxpayers could often extract accumulated corporate value in dividend form in a way that effectively converted capital gains into tax-free or low-tax distributions. The intercorporate dividend deduction, while sound in principle, created opportunities for value extraction that Parliament had not intended.

In that environment, practitioners needed a way to distinguish between dividends that merely distributed income already reflected in share value and dividends that stripped out inherent capital gain. The legislation, however, offered little guidance on how to perform that distinction in practice.

The absence of statutory guidance

Early versions of section 55 were broad and conceptually framed, but thin on mechanics. They articulated outcomes—preventing surplus stripping—without prescribing a detailed computational framework. This left advisors, auditors, and administrators searching for workable methods to operationalize the rule.

Into this gap stepped administrative practice. CRA (and its predecessors) needed a way to review files consistently. Practitioners needed a way to advise clients and prepare computations. Over time, a set of conventions emerged that attempted to approximate what income could safely be distributed without triggering section 55. These conventions eventually coalesced into what became known as the Robertson Rules.

Why Robertson’s Rules emerged

The appeal of the Robertson Rules was pragmatic. They offered a disciplined, spreadsheet-driven approach that appeared to reconcile tax income, accounting surplus, and economic intuition. By adjusting taxable income for items such as taxes paid, dividends paid, and certain non-deductible expenses, the rules aimed to approximate “safe income on hand”—that is, income still available to support share value.

For a time, this approach filled a practical need. It provided consistency in an area where the statute was sparse. It allowed advisors to communicate with CRA using a shared vocabulary. And it reduced uncertainty in routine planning scenarios.

The problem was not that the Robertson Rules existed. The problem was that, over time, they began to be treated as if they were the law.

 

1. Continued practical utility

Despite their limitations, the Robertson Rules have not lost all relevance. Used properly, they still serve a purpose in modern practice.

A risk-management starting point

At their best, the Robertson Rules function as a starting point. They force practitioners to confront the fact that not all taxable income can be safely distributed, and that certain reductions—such as taxes and dividends previously paid—matter when assessing how much income remains to support share value.

As a preliminary screening tool, this discipline remains useful. A computation that ignores taxes paid or assumes all historical income remains intact is unlikely to be credible. The Robertson framework can help identify obvious overstatements of safe income before a deeper statutory analysis is undertaken.

Internal computation discipline

The rules also impose internal rigor. They encourage year-by-year tracking, reconciliation to tax returns, and consistency across periods. In firms where multiple people touch a file over time, that discipline reduces the risk of ad hoc adjustments and undocumented assumptions.

This internal value should not be underestimated. Many section 55 failures stem not from aggressive planning, but from drift—safe income schedules that evolve informally, lose their connection to source data, and eventually become unreliable.

Audit communication shorthand

Finally, the Robertson Rules persist as a form of audit shorthand. Many CRA auditors are familiar with the framework, even if they no longer treat it as determinative. Presenting a computation in a recognizable format can facilitate dialogue, particularly at early stages of an audit.

That said, this utility is conditional. The rules may open the conversation, but they cannot end it. When the analysis turns to paragraph 55(2.1)(c)—and it will—the statutory contribution test must take over.

 

1. Legal limits of the Robertson Rules

The point at which the Robertson Rules cease to be helpful is the point at which they are mistaken for authority. The courts have been clear: administrative heuristics cannot override statutory text.

No statutory footing

The most fundamental limitation is that the Robertson Rules have no statutory footing. They do not appear in the Act. They are not incorporated by reference. They are not endorsed by Parliament.

This matters because section 55 is a recharacterization provision. When it applies, it fundamentally alters the tax treatment of an amount. In that context, courts demand clear legislative authorization. Adjustments that are not grounded in the statute are not merely optional; they are impermissible.

This principle was crystallized in the jurisprudence culminating in Kruco, where the courts rejected attempts to import administrative adjustments—such as hypothetical deductions or retroactive recalculations—into the safe income computation. The message was unambiguous: if the Act does not authorize the adjustment, it does not belong in the calculation.

Over-reliance on accounting constructs

A second legal weakness of the Robertson Rules is their reliance on accounting concepts. Adjustments based on retained earnings, reserves, accruals, or unclaimed depreciation may make sense from a financial reporting perspective, but they are not necessarily aligned with the tax policy embedded in section 55.

The statute does not ask whether accounting profit exists. It asks whether income earned or realized can reasonably be considered to contribute to capital gain. That is a different inquiry, governed by tax law, not accounting standards.

Courts have consistently resisted attempts to substitute accounting logic for statutory interpretation. While accounting data may inform the analysis, it cannot dictate the outcome. To the extent the Robertson Rules blur that distinction, they invite error.

Creation of “phantom income”

Perhaps the most damaging consequence of uncritical reliance on the Robertson Rules is the creation of what is often referred to as phantom income. This occurs when adjustments are made to taxable income to reflect amounts that were never economically realized or that no longer exist in any meaningful sense.

Examples include adding back denied deductions that did not preserve value, adjusting for unclaimed capital cost allowance as if it were income, or treating notional accruals as if they supported share value. These adjustments may inflate a safe income figure, but they do so by severing the link between income and capital gain.

The courts have been explicit in rejecting this approach. Safe income is not a tool for manufacturing distributable amounts. It is a limitation designed to prevent the distribution of amounts that do not reflect income contributing to gain. Phantom income undermines that purpose and fails the statutory test.

 

1. The correct modern role of the Robertson Rules

Given these limits, what role—if any—should the Robertson Rules play in 2026?

Robertson as heuristic, not authority

The correct answer is that the Robertson Rules should be treated as a heuristic. They can help organize thinking, impose discipline, and identify obvious issues. They cannot determine the outcome of a section 55 analysis.

A heuristic guides judgment; it does not replace it. Used this way, the Robertson framework can coexist with the statutory analysis required by paragraph 55(2.1)(c). Used incorrectly, it becomes a substitute for that analysis—and a liability.

Why courts tolerate starting points but not conclusions

Courts understand that complex tax provisions require practical tools. They are not hostile to methodologies that help practitioners navigate difficult terrain. What courts will not tolerate is the elevation of those methodologies to the status of law.

This distinction explains why courts often allow Robertson-style computations as starting points but insist on moving beyond them. A taxpayer may begin with a structured computation, but must ultimately demonstrate, with reference to the statute, that the income relied upon reasonably contributes to the capital gain on the shares.

In other words, the Robertson Rules may help you get to the door. They will not get you through it.

 

Conclusion to this section

The enduring influence of the Robertson Rules reflects their historical utility, not their legal authority. In 2026, their proper role is modest but not nonexistent: they are a tool for organization and communication, not a substitute for statutory interpretation.

Practitioners who continue to treat the Robertson Rules as determinative do so at their peril. Those who understand their limits—and who anchor their analysis in the statutory contribution test—can still use them effectively as part of a broader, defensible safe income framework.

The next sections will trace how the courts drew these boundaries explicitly, and how CRA has adapted its audit posture in response—often imperfectly, but always with an eye toward the same statutory question: what income actually explains the capital gain being extracted?

 

1. The Case Law Inflection Point

(Integrating Kruco scholarship explicitly)

For practitioners trying to rank “Section 55 safe income” and actually win files in real life, the hardest part is not explaining the statute. It is explaining why the same two words—safe income—mean something materially different depending on whether you are talking to (i) a deal lawyer, (ii) a tax preparer, (iii) a CRA auditor, or (iv) the Federal Court of Appeal.

Canadian jurisprudence has forced a reset. The courts have repeatedly pulled the analysis back to first principles: subsection 55(2) is an anti-surplus-stripping rule, and the carve-out for income is not a vibe, a spreadsheet convention, or a CRA memo—it is a statutory construct anchored in the contribution test in paragraph 55(2.1)(c).

This section maps the inflection point: the early cases that framed the policy and mechanics, followed by Brelco and Kruco, which sharpened the legal meaning of safe income and—crucially—put enforceable limits on “administrative creativity.”

 

1. Early judicial framing (what these cases decided, and what they did not)

The early section 55 cases are frequently cited, frequently summarized, and frequently misused. Their enduring value is not that they provide a universal safe income template. Their value is that they establish the conceptual guardrails that still govern modern disputes under 55(2), including today’s “safe income on hand” audit fights.

454538 Ontario Ltd. (TCC)

This case is often referenced (including in CRA commentary) for a foundational proposition: safe income begins with tax income, not accounting income. That is, “income earned or realized” is rooted in the Act’s computation of income (Division B, Part I), and any departures require statutory authorization. This framing is consistent with CRA’s own historical summaries of safe income as starting from net income for tax purposes and then applying legislated adjustments.

What 454538 did (as it is commonly applied in modern safe income disputes) is reinforce that safe income is not a free-form economic concept. You do not get to invent add-backs, remove deductions because they “feel wrong,” or import financial statement logic unless the statute tells you to.

What it did not do is eliminate the economic overlay entirely. Even when safe income is computed as tax income, 55(2.1)(c) still asks whether that income can reasonably be considered to contribute to the capital gain—an inquiry that can require facts about what happened to value over time.

Placer Dome (FCA) — purpose and anti-avoidance framing

In the section 55 context, Placer Dome is frequently cited for its clear articulation of what subsection 55(2) is doing at a policy level: preventing the conversion of what is economically capital gain into a tax-free intercorporate dividend. That framing matters because it explains why “deductible” under 112(1) is never the end of the story, and why CRA auditors gravitate toward the “surplus extraction” narrative even when a taxpayer has a thick binder of safe income schedules.

What Placer Dome did is provide a crisp anti-avoidance lens: if the dividend is being used to strip value that is not attributable to after-tax income, section 55 is designed to intervene.

What it did not do is define safe income numerically or endorse any particular computational approach. Placer Dome is the “why,” not the “how.”

Nassau Walnut (FCA) — safe income allocation and 55(5)(f) mechanics

Nassau Walnut is the leading authority most often pulled into two practical disputes:

1. How safe income is allocated across shares, particularly where shares of the same class may not have identical holding histories; and
2. How paragraph 55(5)(f) operates procedurally (designation timing / mechanics) in the real world of assessments and reassessments.

The case is widely cited in professional commentary for confirming that 55(5)(f) is not a “true election” in the ordinary sense and that taxpayers can, in appropriate circumstances, still access the relief mechanism when 55(2) is invoked by the Minister—even where the designation mechanics were not perfectly executed at filing.

It is also used in discussions about whether safe income must be prorated across shares. Nassau Walnut is regularly referenced in that allocation debate (including in professional summaries addressing share-by-share safe income variability).

What Nassau Walnut did is make section 55 more operational: it addressed how the statute functions in practice when the Minister asserts 55(2), and how the “safe income portion” interacts with the deeming and designation machinery.

What it did not do is bless “spreadsheet folklore” such as phantom income or implied adjustments. It is a mechanics case, not an invitation to depart from the Act’s computation of income.

Gestion Jean-Paul Champagne (and the “on hand” concept’s judicial traction)

Gestion Jean-Paul Champagne is frequently cited for the proposition that safe income must be meaningfully connected to value—that income is not protective if it is no longer available in a way that can reasonably be considered to contribute to the capital gain that the dividend is reducing. This “on hand” idea is one reason why CRA auditors now press so aggressively on cash outflows, value erosion, taxes, and prior dividends.

You can see this logic echoed in later appellate language emphasizing the need to inquire whether income was “kept on hand” or remained disposable to fund the dividend—often illustrated by the statement that taxes and dividends paid out of that income must generally be extracted from safe income.

What Champagne did (as it is used today) is strengthen the legitimacy of an economic overlay—without converting safe income into an accounting surplus concept.

What it did not do is authorize CRA to redefine safe income itself. The overlay operates on the contribution question; it does not permit “phantom adjustments” to the underlying tax-income computation.

 

 

1. Brelco and Kruco — the jurisprudential hardening of safe income

The early cases set the frame. Brelco and Kruco are where the courts begin to draw hard boundaries: safe income is anchored in tax income, and adjustments are not a matter of administrative taste.

1. Safe income is tax income (not retained earnings)

Kruco is regularly cited for a principle that has become the center of modern safe income disputes: safe income is determined by reference to the Act’s computation of income, not by financial statement surplus and not by reconstructed “economic income” unless the statute explicitly compels those adjustments.

That is why professional guidance (and CRA’s own historical technical notes) treat net income for tax purposes as the starting point.

This is also why the “safe income on hand” debate must be carefully handled. The courts do not deny that contribution requires factual analysis. They deny that this gives anyone permission to rewrite the tax base.

2. Adjustments must be legislated (subsection 55(5) is not optional)

The practical significance of Kruco-era jurisprudence is that it disciplined the computation. Courts were unwilling to accept implied adjustments that are not grounded in the Act, particularly where those adjustments materially change the safe income available to support a dividend.

This legal constraint matters in 2026 because CRA audit positions often slide between two different claims:

• a statutory claim about what “income earned or realized” is; and
• an evidentiary claim about whether that income can reasonably be considered to contribute to the capital gain.

Kruco forces practitioners to separate those questions cleanly. If you do not, you invite the auditor to treat every economic objection as a license to re-engineer the safe income number itself.

3. Rejection of phantom income, unclaimed CCA, and retroactive administrative policy

This is where Kruco (and the cluster of jurisprudence around it) remains a live grenade.

In safe income fights, CRA has historically attempted to treat certain items as though they must reduce safe income even where the Act did not require it, or treat certain “forgone deductions” (such as unclaimed CCA) as though they should be retroactively inserted into the computation. Taxpayers, for their part, have sometimes tried to inflate safe income using notional add-backs that do not reflect anything that actually contributed to value.

The courts’ response, distilled, is this: you cannot manufacture safe income. And you cannot erase it through administrative afterthought. Where Parliament wanted adjustments, it legislated them. Where it did not, you do not get to “policy your way” into a different number.

This is precisely why Kruco is still cited in modern appellate discussion of section 55 principles: it remains part of the doctrinal backbone for why section 55 protects dividends attributable to already-taxed corporate income, and why the analysis must be anchored in the statute and the contribution concept.

Brelco — why it matters even when your file is not “about foreign affiliates”

Brelco is often referenced in the safe income literature for the consolidation question (including how negative safe income in one pocket may or may not offset positive safe income in another, depending on the statutory and factual context). Even where your client has no foreign affiliates, Brelco matters because it exemplifies something CRA auditors do constantly: treat safe income as a consolidated, economic metric and then argue about “reasonableness” of the Minister’s approach.

The jurisprudence around Brelco is part of the broader story: courts will allow practical methods, but they will still ask whether the approach is faithful to the statute and the relevant factual linkage to value.

 

1. Why Kruco still matters in 2026

A lot of practitioners treat Kruco as “old law” because the CRA has continued to publish evolving administrative commentary and because the 2015 amendments to section 55 changed the structure of the rule. That is a mistake. Kruco’s importance in 2026 is not about nostalgia. It is about binding constraints that continue to govern how safe income disputes must be argued and won.

Binding legal principles that still run the table

Three Kruco-adjacent principles are still foundational in 2026:

First, safe income is rooted in tax income, not accounting surplus.

Second, adjustments must be legislated—the statute is the source of authority, not administrative habit.

Third, the contribution inquiry under 55(2.1)(c) is real, but it is not a blank cheque. It is an attribution question: whether the income can reasonably be considered to contribute to the capital gain being reduced.

Limits on CRA discretion (and how that shows up in audits)

CRA can select issues, ask questions, and challenge assumptions. CRA cannot redefine the legal meaning of safe income by administrative preference. That is the practical value of Kruco: it gives you a principled basis to push back when the audit drifts from “prove contribution” into “recompute the tax base.”

You can also see how post-Kruco debates became institutionalized in CRA technical commentary and third-party professional summaries describing CRA’s evolving position on phantom income and safe income on hand—often noting that CRA’s views have shifted materially over time.

The foundation for a defensible computation

In 2026, a computation that “passes CRA scrutiny” is not the one that agrees with the auditor’s first spreadsheet. It is the one that can withstand:

1. statutory cross-examination (start with tax income; only legislated adjustments);
2. a contribution narrative tied to 55(2.1)(c); and
3. clear documentation showing what happened to value (taxes, dividends, distributions, real value erosion), without collapsing into phantom adjustments.

That is why Kruco still matters. It is the doctrinal bridge between “law on the books” and “audit reality”—and it is the foundation for the practical template you want to include later in this flagship piece.

 

Quick Access to Cases

454538 Ontario Ltd. (safe income):

https://www.canlii.org/en/#search/type=decision&text=454538%20Ontario%20safe%20income

 

Canada v. Placer Dome Inc., 1996 CanLII 4094 (FCA):

https://www.canlii.org/en/#search/type=decision&text=Canada%20v.%20Placer%20Dome%201996%20CanLII%204094

 

Canada v. Nassau Walnut Investments Inc., 1996 CanLII 4097 (FCA):

https://www.canlii.org/en/#search/type=decision&text=Nassau%20Walnut%20Investments%201996%20CanLII%204097

 

Gestion Jean-Paul Champagne (section 55 safe income):

https://www.canlii.org/en/#search/type=decision&text=Gestion%20Jean-Paul%20Champagne%20section%2055%20safe%20income

 

Canada v. Brelco Drilling Ltd., 1999 CanLII 8151 (FCA):

https://www.canlii.org/en/#search/type=decision&text=Canada%20v.%20Brelco%20Drilling%201999%20CanLII%208151

 

Canada v. Kruco Inc., 2003 FCA 284:

https://www.canlii.org/en/#search/type=decision&text=Canada%20v.%20Kruco%202003%20FCA%20284

 

1. CRA’s Post-Kruco Evolution and Current Audit Position (2025–2026)

If section 55 safe income were merely a statutory interpretation problem, the topic would be difficult but stable. The real challenge—and the reason “55(2)” remains a high-risk file type in 2026—is that CRA’s administrative approach to safe income has not been linear. It has been cyclical. CRA has advanced positions, retreated after adverse jurisprudence, reasserted similar themes using different statutory language, and—more recently—shifted its audit narrative toward the words Parliament actually used in paragraph 55(2.1)(c): income that can reasonably be considered to contribute to the capital gain on the shares immediately before the dividend.

This section is therefore not a “what CRA thinks” overview in the abstract. It is an audit reality map: how CRA’s post-Kruco positions evolved, what CRA focuses on now, how the modern two-step lens operates, and where disputes predictably arise—especially around non-deductible expenses, phantom income, and the survival of “safe income on hand” as a practical (but not statutory) concept.

1. CRA’s cyclical responses after Kruco

A defensible section 55 safe income position in 2026 requires understanding the administrative history because many CRA audit objections are still drawn from older technical news positions—sometimes explicitly, sometimes implicitly—especially when auditors revert to shorthand like “safe income on hand must be reduced by X.” That history is also why many legacy safe income spreadsheets are mismatched to CRA’s current contribution-based framing.

ITTN 33 and ITTN 34 — the retreat, then the refinement

CRA’s archived Income Tax Technical News No. 33 and No. 34 are important not because they are “law,” but because they show CRA’s historical thinking on safe income on hand and phantom income. Both documents speak to CRA’s prior position that safe income on hand should be reduced by phantom income—income that is not represented by an actual receipt of funds.

ITTN 34 specifically frames CRA’s historical approach as reducing safe income by “any actual or potential disbursement or outlay” in the holding period that had not otherwise been deducted in net income for tax purposes, and also reducing safe income on hand for phantom income.

For practitioners, two points matter here:

First, these technical news releases illustrate CRA’s instinct to move beyond statutory safe income and toward an “economic survivability” model. Second, they show CRA’s long-standing discomfort with computations that appear to permit a dividend out of income that does not exist economically.

But this is also where Kruco matters. The Kruco line of jurisprudence constrained CRA’s ability to treat administrative reductions as if they were statutory adjustments. That tension—economic survivability versus statutory authority—has defined CRA safe income controversies ever since.

ITTN 37 — reassertion through “on hand” and non-deductible expenses

CRA’s Income Tax Technical News No. 37 is the pivot point for many modern disputes. ITTN 37 reflects CRA’s view that the amounts that must reduce a corporation’s “safe income on hand” are not limited only to taxes and dividends.

That position—expanded reductions beyond taxes and dividends—became the administrative foundation for CRA challenges involving non-deductible expenses, disbursements, and other cash outflows that were not reflected as deductions in the corporation’s net income for tax purposes.

This is precisely where CRA’s cyclical posture becomes visible:

• After jurisprudence constrained “phantom adjustments” to statutory safe income, CRA continued to press reductions at the “on hand” stage.
• CRA increasingly reframed the debate as one of “reasonable contribution to capital gain,” rather than a debate about whether CRA could rewrite safe income itself.

In other words, CRA did not abandon its economic survivability instinct. It changed how it argued for it.

Selective reliance on Gestion Jean-Paul Champagne

CRA has also relied—selectively—on case law supporting an “on hand” lens, often citing Gestion Jean-Paul Champagne for the proposition that safe income must remain available in a way that contributes to value. Practically, this shows up in audits as a narrative: income that has been consumed by taxes, dividends, or other value-eroding outflows cannot reasonably be said to support the accrued capital gain.

The key word is “selective.” CRA may invoke the economic overlay that Champagne represents, while simultaneously asserting positions that courts have historically resisted when they cross into statutory re-engineering. This is not merely academic. It is exactly how section 55 safe income disputes are litigated: the taxpayer pushes statute, CRA pushes contribution and survivability, and the file turns on whether the taxpayer can demonstrate the link between income and accrued gain using evidence rather than proxies.

1. CRA’s current audit lens: the modern two-step approach

In 2025–2026, CRA’s practical approach to safe income can be summarized as a two-step audit lens:

First, compute statutory safe income. Second, determine what portion of that statutory safe income can reasonably be considered to contribute to the capital gain on the shares immediately before the dividend. This framing is explicitly tied to paragraph 55(2.1)(c) and CRA’s public technical presentations and papers on subsection 55(2) and safe income.

Step 1: Statutory safe income

CRA’s starting point remains that annual safe income generally begins with net income for tax purposes, adjusted by the relevant parts of subsection 55(5). This is consistent with both CRA historical positions and the way CRA continues to describe the computation in various interpretive materials.

In audits, CRA will typically expect that a taxpayer can:

• Tie each year’s starting point to filed T2 information.
• Identify the relevant holding period and the safe-income determination time.
• Segregate safe income on a corporation-by-corporation and share-by-share basis.

If your computation starts with retained earnings, an accounting trial balance, or a consolidated management statement, CRA will typically treat that as an unsupported proxy rather than a safe income computation.

Step 2: Economic contribution to capital gain

Where modern CRA audits become more aggressive is at Step 2. CRA now places substantial weight on whether income can reasonably be considered to contribute to the capital gain on the shares immediately before the dividend. CRA’s public “safe income update” materials emphasize that valuation timing and the determination of FMV must be made prior to the declaration/payment of the dividend, consistent with the statutory test.

This is also where CRA’s current view on phantom income and related issues becomes relevant. In CRA’s 2023–2025 updates as summarized by major firms and professional bodies, CRA has taken the position that certain forms of phantom income should be excluded from safe income because they cannot reasonably contribute to capital gain in a way that supports a dividend.

Whether one agrees with the breadth of CRA’s position, the audit implication is clear: CRA will pressure taxpayers to demonstrate that the income being claimed as “safe” is not merely taxable income on paper, but income that actually explains the accrued gain. Your computation must therefore be paired with an evidentiary narrative.

1. What CRA focuses on in practice (audit reality)

In section 55 safe income audits, CRA’s objections are remarkably consistent. They generally cluster around four themes: cash outflows, value erosion, timing mismatches, and share-specific analysis.

Cash outflows

CRA focuses intensely on cash outflows because cash outflows are the easiest way to argue that income no longer supports share value. Taxes paid, dividends paid, and non-deductible disbursements are the common categories CRA emphasizes. This audit focus is reflected in CRA’s historic technical news and in more current CRA roundtable commentary addressing what constitutes non-deductible expenses that reduce safe income contributing to gain.

The practical lesson is that a safe income computation that does not track cash-based reductions (even if it tracks tax income perfectly) will look incomplete to CRA.

Value erosion

CRA also focuses on value erosion—losses, write-downs, unsuccessful investments, and expenditures that do not create enduring asset value. CRA’s framing is that income cannot contribute to capital gain if it has been consumed without leaving value behind. This is a factual inquiry, and it is why contemporaneous documentation matters: the reason for the outlay, the use of funds, and the connection to asset value are often determinative.

Timing mismatches

Timing mismatches are among the most common “gotchas” in section 55 audits. CRA will test whether income relied upon to support a dividend was earned during the relevant period, and whether it existed at the safe-income determination time. This includes scrutiny of stub periods, multi-step series, and the sequencing of dividends and reorganizations—especially where a dividend is paid as part of a pipeline, butterfly, or pre-sale extraction.

CRA’s public materials underscore that the FMV and accrued gain analysis is tied to the time immediately before the dividend.

Share-specific analysis

Finally, CRA often insists on share-specific safe income analysis. This is particularly important in estate freezes, share reclassifications, and reorganizations where different share classes represent different entitlements and different inherent gains. CRA’s concern is that a taxpayer is using corporate-level income as a blanket justification for dividends on shares that did not economically benefit from that income.

This is why safe income is not merely corporation-specific. It is share-specific. And when that share specificity is ignored, CRA often treats the computation as conceptually flawed.

1. Non-deductible expenses — the narrow, defensible view

Non-deductible expenses are where section 55 safe income disputes become most contentious, because the topic sits at the intersection of three themes: statutory computation, economic survivability, and CRA’s tendency to generalize.

CRA has provided a general description of non-deductible expenses that it considers will reduce the portion of safe income that can reasonably be considered to contribute to capital gain, notably at the 2021 STEP Roundtable. CRA’s description focuses on non-deductible expenses as cash outflows not deducted in computing net income for tax purposes, with certain exclusions.

For an audit-defensible approach in 2026, the narrow, defensible principle is this:

Non-deductible expenses should reduce safe income only to the extent they represent cash outflows that actually erode economic value and therefore reduce the income that can reasonably be considered to contribute to the capital gain on the shares immediately before the dividend.

That sentence matters because it separates legitimate reductions from overreach.

Only reduce safe income where real economic value is lost

If a non-deductible expenditure is a cash outflow that does not create or preserve asset value and is not otherwise reflected in the tax-income computation, CRA will argue it reduces the portion of safe income that contributes to gain. In many cases, this is defensible: a cash outflow that permanently consumes value is difficult to reconcile with a claim that the same income still supports share value.

No reduction merely because the Act denies a deduction

But the denial of a deduction, by itself, does not prove value erosion. There are many expenses denied under the Act for policy reasons that may nonetheless have created or preserved value, or may be reflected elsewhere in the share value. If an auditor treats “non-deductible” as equivalent to “does not contribute to capital gain,” the analysis has become categorical rather than factual.

CRA’s own framing ties the inquiry to paragraph 55(2.1)(c)—the portion of safe income that can reasonably be considered to contribute to gain. That is a reasonableness test, not a bright-line rule.

Where CRA overreaches in audits

Overreach tends to occur in three patterns:

First, when CRA pushes reductions for items that are not true economic outflows (notional adjustments, reserves, accounting entries). Second, when CRA treats denied deductions as automatically value-eroding without examining facts. Third, when CRA seeks to “fix” perceived inequity by importing phantom deductions or hypothetical adjustments, rather than staying within the statutory framework.

Modern professional commentary reflects that CRA’s positions on phantom income and related concepts have continued to evolve, and that CRA now emphasizes contribution-to-gain reasoning rather than the older mechanical “on hand” computations.

The practical implication for planners is straightforward: you must prepare a safe income analysis that is both numerically correct and narratively coherent. You should assume CRA will test non-deductible expenses not as a tax technicality, but as an evidentiary lever to argue that income did not survive to contribute to gain.

 

VII. How to Calculate Safe Income — A CRA-Defensible Template

A “safe income” computation that survives a section 55 audit is not the one with the most elaborate spreadsheet or the longest list of adjustments. It is the one that can be defended under sustained scrutiny using three pillars: the statute, the facts, and a coherent explanation of contribution to capital gain.

That explanation is no longer optional. CRA has explicitly tied its current audit posture to paragraph 55(2.1)(c)—whether the income relied upon can reasonably be considered to contribute to the accrued capital gain on the shares immediately before the dividend. CRA has also repeatedly stated that safe income on hand must be reduced for real economic outflows, including non-deductible expenses that reduce disposable after-tax income. The unifying theme is contribution to value, not mechanical surplus tracking.

This section sets out a repeatable, five-step calculation framework that can be used in live planning files, internal technical reviews, and audit defence. It is written for tax lawyers, senior CPAs, and tax managers, and is intentionally structured so it can be converted into a signature firm deliverable—whether as a standalone safe income memo, an appendix to a reorganization plan, or a formal response to a CRA information request.

Before turning to the mechanics, one structural point bears emphasis. This framework separates the analysis into three distinct layers:

1. Statutory safe income (what the Act permits you to count);
2. Economic survivability (what remains to support value); and
3. Capital gain reconciliation (what the income actually explains).

Most failed computations collapse these layers into a single spreadsheet. A CRA-defensible computation never does.

 

1. Why most safe income computations fail audit

Most safe income computations fail for predictable reasons that have little to do with arithmetic and everything to do with framing.

First, many analyses begin with the wrong anchor. Retained earnings, accounting surplus, or management equity schedules are often used as starting points because they are readily available. CRA has stated—repeatedly—that retained earnings may be accepted only in very rare cases and only after stringent validation. That alone should signal how risky it is to rely on accounting proxies as the foundation of a section 55 analysis. The statute and the post-Kruco jurisprudence push the analysis back to tax income, not accounting constructs.

Second, practitioners frequently over-adjust for non-cash items. There is a strong temptation to “correct” taxable income into an intuitive economic measure by adding back unclaimed CCA, reversing reserves, or normalizing timing differences. These adjustments often feel reasonable. They are also where the analysis usually collapses. Safe income is a statutory construct, and subsection 55(5) provides a finite list of permitted modifications. When a computation becomes a hybrid of tax law, accounting logic, and intuition, it can no longer be explained coherently—either to CRA or to a court.

Third, and most importantly, many computations fail the capital gain reconciliation. Paragraph 55(2.1)(c) is not satisfied by a number in isolation. The amount of safe income claimed must be plausibly linked to the capital gain that could be realized on the shares immediately before the dividend. If the safe income figure exceeds, or cannot explain, the accrued gain, the computation is conceptually vulnerable even if the arithmetic is flawless.

A CRA-defensible template must therefore do three things at once. It must compute statutory safe income correctly. It must apply an economic filter grounded in actual value erosion, not notional adjustments. And it must reconcile the result to the accrued capital gain at the correct moment in time.

What follows is a five-step method that does exactly that.

 

Step 1: Identify the Relevant Shares and the Relevant Period

A safe income computation is only as strong as its share identification and period definition. In practice, this is where many files fail before the numbers are ever tested.

Begin by identifying the exact shares in respect of which the dividend will be paid or deemed to be paid, including redemptions under subsection 84(3). This requires more than naming the corporation. You must identify the share class, the legal holder, and the economic entitlement attached to those shares. Common shares, fixed-value preferred shares, growth shares, and redeemable or retractable shares do not bear income in the same way. Where there are multiple tranches within a class, that distinction must also be documented.

This share-specific focus is not academic. Safe income is intended to relate to the capital gain on the particular shares being stripped. CRA routinely presses for this level of specificity in audits, especially in estate freezes, internal buyouts, and multi-class capital structures. A corporate-level computation without share-level attribution is increasingly difficult to defend.

Next, determine the safe-income determination time. Paragraph 55(2.1)(c) fixes the inquiry at the moment immediately before the dividend, and the relevant accumulation period ends at that point. Where the dividend forms part of a transaction, event, or series, identifying the correct determination time is critical. Income earned after that time is irrelevant, even if it appears economically proximate.

Then identify the acquisition date of the shares. In simple owner-managed structures this may be incorporation. In reorganizations, it may be a share reclassification date, an estate freeze, or the issuance of new shares for nominal consideration. This matters because CRA has long taken the view that pre-acquisition earnings are embedded in the purchase price and are not “new” safe income for the acquirer. That principle becomes especially important when shares change hands within a corporate group.

Finally, document the impact of rollovers and reorganizations, including sections 51, 85, 86, and 87. These provisions alter share attributes, paid-up capital, adjusted cost base, and sometimes the economic relationship between income and gain. A safe income schedule that ignores the reorganization history will almost always be viewed as non-credible in audit.

A brief word on sales and the LCGE. Safe income is not an LCGE concept, but the two intersect in practice because section 55 issues commonly arise in pre-sale surplus extraction, purification, and internal reorganizations undertaken in anticipation of a share sale. A dividend that triggers subsection 55(2) can derail the intended sale economics entirely. The issue is not that LCGE planning changes safe income; it is that a section 55 failure may prevent you from ever reaching the LCGE outcome you planned for.

 

Step 2: Compute Statutory Safe Income

This step is the law-on-the-books computation. It must be kept clean and conceptually separate from any economic filtering.

Start with the corporation’s net income for tax purposes for each year in the relevant period. In practice, Schedule 1 is the most reliable audit anchor. The computation must be prepared corporation-by-corporation. If a dividend is paid from a subsidiary to a parent, you need the subsidiary’s safe income computation on its own merits. Consolidated or group-level proxies are almost always challenged.

Once the starting point is established, apply only those adjustments that the Act expressly permits. Subsection 55(5) is exhaustive. Each adjustment should be explicitly labelled as a statutory adjustment and tied to the relevant paragraph. If you cannot point to the Act, the adjustment does not belong in the statutory layer.

Equally important are the explicit exclusions. A CRA-defensible computation should affirmatively state that it does not include phantom income, does not insert unclaimed CCA, and does not rely on administrative add-backs that lack legislative footing. This is not stylistic. It signals to the reviewer that you understand the limits of the statutory construct and have respected them.

At the conclusion of Step 2, you should have a year-by-year schedule that a court would recognize: tax income first, legislated modifications second, and nothing else.

 

Step 3: Apply the Economic Filter (Safe Income on Hand)

Only after statutory safe income has been computed do you move to the economic filter—what practitioners shorthand as “safe income on hand.”

The phrase itself does not appear in the Act, but CRA has consistently taken the position that safe income must be reduced to reflect actual economic dissipation. The governing principle is narrow and defensible: reduce safe income only for items that erode real economic value, such that the income can no longer reasonably be considered to contribute to the capital gain on the shares immediately before the dividend.

In practice, the core reductions usually include income taxes paid or payable, dividends paid or payable out of the relevant income stream, realized losses that destroy value, and non-deductible expenses that involve genuine cash outflows. CRA’s STEP Roundtable guidance describes non-deductible expenses for this purpose as cash outflows not deducted in computing net income for tax purposes, subject to exclusions such as capital outlays and principal repayments.

This is where judgment matters most.

Over-reduction occurs when non-cash accounting entries—reserves, accruals, contingencies—are treated as value erosion. Under-reduction occurs when practitioners refuse to recognize obvious economic dissipation simply because the Act denied a deduction or because the item is “below the line.” A disciplined analysis does neither. It asks a single question: did this item permanently consume value that would otherwise support share value?

A defensible schedule should also state clearly what it does not reduce for: notional items, denied deductions without cash impact, and accounting reserves. If CRA wishes to argue those reductions, the burden should fall on CRA to tie them back to real value erosion and the statutory contribution test.

The output of Step 3 is a transparent reconciliation showing how statutory safe income is reduced to the amount that remains economically “on hand.”

 

Step 4: Reconcile to Accrued Capital Gain

This step transforms a calculation into a section 55-ready analysis.

Determine the accrued capital gain on the relevant shares immediately before the dividend. This requires credible support for fair market value and adjusted cost base. A formal valuation report is not always required, but unsupported assumptions are not acceptable. The timing is critical. The valuation must align with the statutory moment specified in paragraph 55(2.1)(c).

Once the accrued gain is established, compare it to the safe income on hand computed in Step 3. The logic is straightforward: the portion of the dividend that can be protected as safe income cannot exceed the amount of income that can reasonably be considered to contribute to that gain. If safe income on hand exceeds the accrued gain, the discrepancy must be explained. Often it signals a share-specific attribution problem or a valuation mismatch.

Conclude this step by stating the cap explicitly: the lesser of the accrued capital gain and the safe income on hand attributable to the shares. That is the maximum dividend amount that is typically defensible under section 55.

 

Step 5: Documentation and Evidentiary Support

Section 55 disputes are rarely lost because the taxpayer cannot calculate. They are lost because the taxpayer cannot explain.

A defensible file includes year-by-year schedules tied to filed returns, a methodology memo written in plain statutory language, and a narrative explaining how income translates into value and how that value manifests as capital gain. Where valuation support is required, it should be proportionate—neither perfunctory nor excessive.

A sentence worth including in any section 55 file because it reflects audit reality is this: CRA disputes are won on explanation, not arithmetic. That sentence aligns the analysis with paragraph 55(2.1)(c) and forces the discussion away from spreadsheet folklore and toward statutory contribution.

 

Sub-Appendix A: Sample computation table

Below is a simple table structure that works well in audits because it separates statutory income from the on-hand filter and then ties the result to accrued gain. Populate it year-by-year and include totals for the relevant period.

Year. Net income for tax purposes. Statutory safe income adjustments under 55(5). Statutory safe income subtotal. Taxes paid/payable. Dividends paid/payable. Cash non-deductible expenses. Realized losses. Safe income on hand subtotal. Notes and support references.

This table is intentionally plain. Its strength is that every line item is either tax-return-anchored or fact-anchored, and each reduction is tied to real value erosion.

This template is designed to support the calculation and documentation of safe income and safe income on hand for purposes of subsection 55(2) of the Income Tax Act, using a methodology that is:

• grounded in the statutory framework of section 55 and subsection 55(5);
• aligned with post-Kruco jurisprudence;
• consistent with CRA’s current audit posture (2025–2026); and
• defensible under audit, objection, or litigation.

This template intentionally separates:

1. Statutory safe income (law on the books),
2. Economic survivability / “on hand” analysis, and
3. Capital gain reconciliation.

These components must not be collapsed into a single schedule. CRA scrutiny is significantly reduced when the analytical layers are kept distinct and explained clearly.

 

Instructions

This template must be completed share-specifically, not merely at the corporate level, where different share classes or tranches exist.

1. Each numerical input must be traceable either to:
   • a filed tax return (Schedule 1, T2 schedules), or
   • a factual document (bank records, dividend resolutions, invoices, valuation support).
2. No adjustment may be made unless it can be:
   • tied to a specific statutory provision (Step 2), or
   • justified as real economic value erosion (Step 3).
3. If a number cannot be explained in words, it does not belong in the schedule.

 

Template Structure Overview

This template consists of four core tables and one narrative schedule:

• Table A-1: Share & Period Identification (Step 1)
• Table A-2: Statutory Safe Income Computation (Step 2)
• Table A-3: Safe Income on Hand (Economic Filter) (Step 3)
• Table A-4: Accrued Capital Gain Reconciliation (Step 4)
• Schedule A-5: Narrative & Evidentiary Support (Step 5)

Table A-1

Share Identification & Relevant Period**

(Complete once per share class or tranche)

Item	Description
Corporation
Share class
Legal holder
Economic entitlement
Acquisition date
Method of acquisition	(e.g., incorporation, s.85, s.86, freeze)
Safe-income determination time
Dividend / redemption date
Part of transaction / series?	Yes / No (explain)
Relevant accumulation period	From ___ to ___

Practitioner note:
This table anchors the entire analysis. If CRA challenges the period or the shares, the computation fails regardless of accuracy elsewhere.

 

Table A-2

Statutory Safe Income Computation (s.55(5))**

Instruction:
Start with net income for tax purposes for each year in the relevant period.
Adjust only as expressly permitted by subsection 55(5).
Do not apply economic or “on hand” concepts in this table.

Tax Year	Net income for tax purposes (Sched. 1)	s.55(5) adjustments (specify paragraph)	Statutory safe income for year	Cumulative statutory safe income	Notes / references
20XX
20XX
20XX
Total

Explicit exclusions (state affirmatively):

• No phantom income included
• No unclaimed CCA inserted
• No administrative or accounting add-backs
• No retained earnings proxies used

Practitioner note:
This table should be capable of standing alone before a court. If an adjustment cannot be tied to subsection 55(5), it does not belong here.

 

Table A-3

Safe Income on Hand — Economic Survivability Filter**

Instruction:
Reduce statutory safe income only for items that represent actual economic value erosion such that the income can no longer reasonably be considered to contribute to the capital gain on the shares.

 

 

 

 

 

Tax Year	Statutory safe income brought forward	Income taxes paid / payable	Dividends paid / payable	Cash non-deductible expenses	Realized losses	Safe income on hand for year	Cumulative safe income on hand	Notes / factual support
20XX
20XX
20XX
Total

Do NOT reduce safe income on hand for:

• accounting reserves or provisions
• accrued but unpaid notional items
• denied deductions without cash outflow
• capital expenditures (asset substitution)
• loan principal repayments

Practitioner note:
Every reduction must answer one question:

Did this item permanently consume value that would otherwise support share value?

 

**Table A-4

Accrued Capital Gain Reconciliation**

 

Item	Amount	Notes / valuation support
Fair market value of shares immediately before dividend
Adjusted cost base of shares
Accrued capital gain
Safe income on hand attributable to shares
Maximum defensible dividend (lesser of above)

Practitioner note:
If safe income on hand exceeds the accrued gain, explain why.
Unexplained excesses are a common audit trigger.

 

Schedule A-5

Narrative Explanation & Evidentiary Support**

Purpose of dividend / redemption:
[Explain commercial context]

Explanation of contribution to capital gain:
[Explain how the income relied upon reasonably contributed to share value]

Summary of key assumptions:

• valuation timing
• share entitlements
• treatment of non-deductible expenses

Supporting documents on file:

• T2 returns and Schedule 1s
• dividend resolutions
• bank statements
• valuation reports or internal valuation memo
• reorganization documents (if applicable)

Audit posture statement:

This computation is prepared in accordance with subsection 55(5) and paragraph 55(2.1)(c) of the Income Tax Act and reflects only income that can reasonably be considered to contribute to the accrued capital gain on the shares immediately before the dividend.

 

How CRA Could Use This Template in Audit

When prepared properly, this template forces CRA to engage on law and facts, not folklore:

• CRA cannot challenge Step 2 without identifying a statutory error.
• CRA cannot challenge Step 3 without proving real economic value erosion.
• CRA cannot challenge Step 4 without disputing valuation or ACB.

This is precisely why this structure is effective: it allocates the burden of argument back to CRA, where the statute intends it to be.

 

 

Sub-Appendix B: Audit request checklist

CRA information requests typically converge on the same core items in section 55 safe income audits. A defensible file anticipates these requests.

Your checklist should include: T2 returns and Schedule 1 for each relevant year; corporate minute book extracts supporting dividends and redemptions; share capital history including rollovers and reorganizations; paid-up capital continuity; adjusted cost base computations; details of taxes paid; details of dividends paid; details supporting non-deductible expenses treated as reductions; documentation of significant cash outflows and value-eroding events; and FMV support immediately before the dividend.

You do not need to hand CRA a narrative they can weaponize. But you do need to be able to provide clean source documents that support your computation methodology and your capital gain reconciliation.

Purpose and Use

CRA information requests in section 55 audits are highly patterned. While the phrasing may vary, the underlying requests almost always seek to test three things:

1. Statutory safe income (accuracy and legislative footing),
2. Economic survivability (what income remains “on hand”), and
3. Contribution to capital gain at the correct moment in time.

This checklist is designed to anticipate CRA’s requests, ensure complete documentary readiness, and control the order and scope of disclosure. It supports a disciplined response strategy: provide clean source documents, tie them to the computation, and avoid unnecessary narrative or opinion unless specifically requested.

 

How to Use This Checklist

• Lawyers: Use this checklist to scope the file, confirm readiness, and set boundaries on disclosure.
• Tax Managers: Use this checklist to assemble documents, cross-reference schedules, and prepare a defensible response package.
• Timing: Complete this checklist before any substantive response is sent to CRA.

 

**Table B-1

Core CRA Audit Requests — Mandatory Source Documents**

(These items are almost always requested explicitly or implicitly)

Category	Document / Schedule	Years / Period	Source	Status	Notes
Tax filings	T2 returns	All relevant years	Filed returns	☐	Include notices of assessment
Tax filings	Schedule 1 – Net income for tax purposes	Same years	T2 schedules	☐	Anchor for statutory safe income
Corporate records	Articles, bylaws	N/A	Minute book	☐	Confirm share rights
Corporate records	Directors’ resolutions	Relevant years	Minute book	☐	Dividends / redemptions
Corporate records	Share registers	Entire holding period	Minute book	☐	Share-specific analysis
Capital history	Share issuance / reclassification docs	Relevant events	Legal docs	☐	s.51 / 85 / 86 / 87
Capital history	Paid-up capital continuity	Relevant years	Internal schedule	☐	Reconcile to filings
Tax attributes	Adjusted cost base computations	At dividend time	Internal schedule	☐	Share-specific
Cash flows	Taxes paid / payable	Relevant years	CRA / GL	☐	Link to Step 3
Cash flows	Dividends paid / payable	Relevant years	Resolutions / GL	☐	Timing matters
Expenses	Non-deductible expenses (cash)	Relevant years	GL / invoices	☐	Only if relied upon
Losses	Realized losses	Relevant years	Tax schedules	☐	Capital vs non-capital
Valuation	FMV immediately before dividend	At SIDT	Valuation memo/report	☐	Proportionate support

Execution note:
Each document should be tagged to the relevant table in Sub-Appendix A (A-1 to A-4). If a document does not support a specific line item, question whether it should be included.

 

**Table B-2

Safe Income Computation Support — Cross-Reference Matrix**

(Ensures every number can be traced)

Safe Income Step	Line Item	Supporting Document(s)	Location / Reference
Step 2	Net income for tax purposes	Schedule 1
Step 2	s.55(5) adjustment	Statutory citation + schedule
Step 3	Income taxes paid	CRA statements / GL
Step 3	Dividends paid	Resolutions + GL
Step 3	Non-deductible cash expense	Invoice / proof of payment
Step 4	FMV	Valuation memo/report
Step 4	ACB	ACB schedule

Execution note:
If a line item cannot be cross-referenced to a primary document, it should be removed or revised before responding to CRA.

 

**Table B-3

Transaction & Series Documentation (Where Applicable)**

(Often requested once CRA frames a “series” argument)

Item	Description	Document	Status
Transaction overview	Description of steps	Internal memo	☐
Series identification	Event / transaction linkage	Legal docs	☐
Timing map	Dates of steps	Timeline	☐
Inter-entity flows	Cash / dividends	Bank records	☐
Post-transaction state	Share ownership	Share register	☐

Lawyer note:
Provide factual documentation only. Avoid interpretive memos unless specifically requested.

 

**Table B-4

Value Erosion & Economic Survivability Support**

(Supports Step 3 — “on hand” analysis)

Item	Nature of Outflow	Amount	Evidence	Value Impact Explanation
Taxes	Corporate income tax		CRA / GL	Permanent cash outflow
Dividend	Inter-corporate dividend		Resolution / GL	Value removed
Expense	Non-deductible cash item		Invoice	Value consumed
Loss	Realized loss		Tax schedule	Asset value destroyed

Execution note:
Do not include non-cash accounting items (reserves, accruals) unless CRA explicitly requests them—and even then, provide separately with reservations.

 

**Table B-5

Valuation & Capital Gain Support**

(Focused strictly on paragraph 55(2.1)(c))

Item	Amount	Support	Notes
FMV of shares		Valuation	As at immediately before dividend
ACB		ACB schedule	Share-specific
Accrued capital gain		Computation
Safe income on hand		Sub-Appendix A
Maximum defensible dividend		Comparison

Execution note:
Avoid debating valuation methodology in the initial response. Provide support sufficient to demonstrate reasonableness and timing.

 

What NOT to Provide Unless Requested

A defensible section 55 response is not exhaustive disclosure. Do not provide:

• legal opinions or internal deliberations;
• planning emails or draft memos;
• alternative computations not relied upon;
• narrative explanations beyond what is necessary to connect documents to numbers.

The objective is support, not persuasion.

 

Response Strategy Guidance

1. Stage disclosure: Provide core documents first; respond to follow-ups deliberately.
2. Anchor every response: Tie documents back to Sub-Appendix A tables.
3. Preserve flexibility: Avoid committing to interpretations unless asked.
4. Document intent internally: Keep internal notes separate from CRA submissions.

 

Practitioner Closing Note

A section 55 audit is rarely won by volume. It is won by structure, traceability, and restraint. This checklist ensures that when CRA asks the inevitable questions, the answers are already organized, defensible, and anchored in law and fact—without volunteering material that creates unnecessary risk.

 

 

 

Sub-Appendix C: Common CRA challenges and defensible responses

Three objections are common.

The first is “your safe income is just retained earnings.” The defensible response is that safe income is computed from tax income, with legislated adjustments, and that retained earnings is at most a secondary reasonableness check, not the computation. CRA itself has acknowledged retained earnings is only rarely acceptable as a proxy and only after stringent validation.

The second is “you did not reduce for non-deductible expenses.” The defensible response is that reductions were applied only for true cash outflows that erode value, consistent with CRA’s own description of non-deductible expenses as cash outflows not deducted in computing net income for tax purposes (with exclusions), and consistent with the statutory contribution test.

The third is “your income does not contribute to the gain.” The defensible response is Step 4: the computation is reconciled to accrued capital gain immediately before the dividend, supported by ACB continuity and FMV evidence, consistent with CRA’s contribution-based framing.

 

 

VIII. Section 55 in Real Transactions

Section 55 problems rarely announce themselves at the planning stage. They surface later—often years later—when the transaction is closed, the professional team has disbanded, and the facts have hardened. By that point, what was once a “structuring issue” becomes a reassessment problem, and what could have been managed with planning becomes irreversible.

This is why section 55 remains one of the most professionally dangerous provisions in the Act for advisors working with family-owned enterprises. It operates quietly in the background of transactions that are otherwise commercial, familiar, and well-trodden. The dividend is deductible. The cash moved. The deal closed. And then, long after the focus has shifted, CRA asks a simple question: what portion of that dividend was actually supported by safe income?

This section walks through the transactions where section 55 issues most commonly arise in practice—not as hypotheticals, but as real planning contexts encountered repeatedly in Canadian private-company work. The objective is not promotion. It is clarity: understanding where section 55 risk lives, why it surfaces late, and why early intervention matters.

 

Internal family buyouts

Internal family buyouts are among the most common—and most misunderstood—contexts in which section 55 applies. These transactions often feel benign. Control remains within the family. The business continues operating. The dividend is paid to facilitate a buyout or equalization. Nothing appears aggressive.

From a section 55 perspective, however, internal buyouts are structurally risky because they often involve dividends or redemptions used to eliminate accrued capital gain on shares without an external realization event.

A typical pattern looks like this. A parent or sibling group holds common shares of an operating company or holding company. One family member is exiting. The remaining family members want to consolidate ownership. Rather than triggering a sale to a third party, the transaction is structured internally: shares are redeemed, dividends are paid, or holding companies are interposed to facilitate the buyout.

From a commercial standpoint, the transaction makes sense. From a tax standpoint, it frequently relies—explicitly or implicitly—on the assumption that accumulated corporate earnings can be distributed tax-free through intercorporate dividends. That assumption is precisely what section 55 interrogates.

The problem is timing. At the time of the buyout, advisors are focused on valuation, fairness between family members, financing mechanics, and emotional dynamics. Safe income is often addressed at a high level, if at all. Years later, CRA examines the dividend and asks whether it reduced the inherent gain on the shares beyond what safe income could support.

At that point, the buyout cannot be unwound. The departing family member is gone. The shares have been cancelled or transferred. The only question left is whether the dividend should be recharacterized under subsection 55(2). This is why internal family buyouts are disproportionately represented in section 55 reassessments.

 

Pipelines and post-mortem planning

Pipelines are a technically sophisticated planning technique, but they are also one of the most scrutinized contexts for section 55 safe income analysis. This is especially true in post-mortem planning, where a deceased shareholder’s shares have an accrued gain and the objective is to extract corporate value without triggering double taxation.

In a pipeline, dividends and share redemptions are often sequenced carefully to transform corporate surplus into capital returns. The planning typically relies on the idea that certain dividends should not be caught by section 55 because they are paid out of safe income or because the structure is designed to produce capital gain treatment elsewhere.

CRA understands pipelines extremely well. In audits, CRA rarely attacks the pipeline concept directly. Instead, it focuses on the dividends within the pipeline, asking whether they reduced the inherent gain on shares in excess of safe income.

What makes pipelines particularly vulnerable is that they often involve multiple steps over time: estate freezes, share subscriptions, redemptions, and wind-ups. The safe-income determination time may differ for each step. Income may have been earned, taxed, distributed, or dissipated between steps. If the safe income narrative is not documented contemporaneously, it becomes very difficult to reconstruct years later.

Post-mortem pipelines add another layer of complexity because the deceased shareholder’s adjusted cost base and the fair market value at death anchor the accrued gain analysis. CRA audits in this space frequently focus on whether income earned after death can reasonably be considered to contribute to the gain that existed immediately before certain dividends were paid.

The common failure point is not aggressive planning. It is assumed continuity—the assumption that income earned “around the same time” supports value, without carefully mapping the income to the gain at the precise statutory moment. Section 55 is unforgiving on this point.

 

Estate freezes

Estate freezes are a cornerstone of succession planning for family-owned enterprises, but they are also a recurring source of section 55 exposure. The risk does not usually arise at the moment of the freeze. It arises later, when dividends are paid on the frozen shares or when the frozen shares are redeemed or reorganized.

In a typical freeze, the growth in the business is shifted to new common shares held by the next generation or a family trust. The original owner retains fixed-value preferred shares. Those preferred shares often carry dividend rights, redemption features, or both.

Years later, dividends are paid on the preferred shares to extract value, fund retirement, or facilitate further reorganization. From a business perspective, this feels like distributing accumulated earnings. From a section 55 perspective, the question is more precise: does the income relied upon actually explain the capital gain embedded in those preferred shares at the time of the dividend?

This is where many freezes run into trouble. The accrued gain on the preferred shares is often fixed at the time of the freeze. Income earned after the freeze may increase the value of the common shares, not the preferred shares. Yet dividends paid on preferred shares are sometimes justified using corporate-level safe income without share-specific analysis.

CRA audits in this area focus heavily on share-specific contribution. If the income did not economically accrue to the frozen shares, CRA will argue it cannot reasonably be considered to contribute to their capital gain. At that point, even a carefully prepared corporate-level safe income schedule may not protect the dividend.

Estate freezes therefore illustrate a broader lesson: section 55 analysis cannot stop at the corporation. It must follow the value to the shares on which the dividend is paid.

 

Pre-sale surplus extraction

Pre-sale surplus extraction is perhaps the most intuitive context for section 55, and yet it is still one of the most common sources of reassessment. The goal is straightforward: remove excess cash or surplus from a corporation before selling shares to a third party, often to improve integration or avoid selling “trapped” cash.

In practice, this often involves intercorporate dividends from an operating company to a holding company, followed by distributions to shareholders. The dividends are deductible under subsection 112(1). The cash moves. The sale closes.

Years later, CRA examines the transaction and focuses on a single question: did the dividend reduce the inherent capital gain on the shares that were sold? If the answer is yes, CRA then asks whether the amount of the dividend exceeded the safe income that could reasonably be considered to contribute to that gain.

The difficulty with pre-sale surplus extraction is that the valuation of the business at sale time often incorporates both operating value and excess cash. If the dividend is paid shortly before the sale, CRA may argue that the cash distributed was already reflected in the share value and that the dividend therefore reduced the capital gain.

This is not a theoretical concern. CRA audit practice in this area is well developed, and the outcomes turn on fine factual distinctions: timing, valuation assumptions, how purchase price was negotiated, and whether the dividend was contemplated as part of the sale process.

What makes these cases particularly painful is that the sale is complete. The purchase price is fixed. The seller cannot retroactively change the structure. Section 55 issues surface only after closing, when the only remaining question is the character of the dividend for tax purposes.

 

Why section 55 issues surface after closing

Across all of these transaction types, a common pattern emerges. Section 55 issues rarely surface at filing time because nothing appears wrong on the face of the return. The dividend is deductible. The capital gain is reported (or not). The numbers reconcile.

CRA reassessments occur later because section 55 is a contextual provision. It requires CRA to reconstruct what happened economically, often years after the fact, and to ask whether the dividend had the prohibited effect. This type of analysis is resource-intensive and typically reserved for audits, not initial assessments.

By the time CRA raises section 55, the transaction is irreversible. Shares have been redeemed. Family members have exited. Estates have been wound up. Sale proceeds have been spent or reinvested. There is no practical remedy other than disputing the reassessment.

This is why section 55 should be treated as a planning-stage risk, not an audit-stage problem. The cost of addressing safe income before a transaction is trivial compared to the cost of defending a reassessment years later.

For family-owned enterprises, lawyers, and senior accounting professionals, this is the practical takeaway: section 55 does not punish bad intentions. It punishes unsupported assumptions. And it does so quietly, long after the opportunity to fix the structure has passed.

The next section draws those lessons together and translates them into clear guidance on when section 55 must be addressed explicitly—and when silence is no longer a defensible strategy.

 

1. Why Family-Owned Enterprises Are Uniquely Exposed

Section 55 risk does not distribute evenly across the Canadian economy. Public companies, institutional investors, and widely held corporate groups tend to encounter section 55 in narrow, well-documented contexts. Family-owned enterprises, by contrast, experience section 55 as a latent risk—one that accumulates quietly over decades and surfaces only when the business reaches a transition point.

This is not because family enterprises are aggressive. It is because their governance, capital structures, and decision-making processes evolve organically rather than transaction-by-transaction. Section 55, however, does not care about intentions or history. It asks a narrow statutory question: whether a dividend has reduced an inherent capital gain beyond what safe income can support.

The features that make family-owned enterprises resilient and enduring are often the same features that make them vulnerable under section 55. Understanding that vulnerability is essential for advisors who work in this space.

 

 

Long holding periods

Family-owned enterprises often hold shares for decades. It is common to see shareholders who acquired their shares on incorporation or in the early growth years of the business and have never disposed of them. Over time, the business evolves: operations expand, assets are acquired and sold, subsidiaries are added, holding companies are interposed, and income is earned, taxed, and reinvested.

From a business perspective, this continuity is a strength. From a section 55 perspective, it creates complexity.

Long holding periods mean that the safe income accumulation period can span many years, sometimes across multiple tax regimes, accounting standards, and business models. Records may be incomplete. Dividend histories may be informal. Corporate reorganizations may have occurred without a contemporaneous safe income analysis.

CRA audits in this context often involve reconstructing a multi-decade history of income, distributions, and value changes. The longer the holding period, the harder it is to demonstrate that the income relied upon can reasonably be considered to contribute to the capital gain on the shares at a specific point in time. Even where the numbers ultimately support the taxpayer, the evidentiary burden is heavier.

This is why section 55 problems in family enterprises tend to surface late: the risk accumulates slowly, and only becomes visible when a transaction forces CRA to look backward.

 

Informal dividend histories

In many family-owned enterprises, dividends are paid informally. Decisions are often made at the board table or around the kitchen table. Dividends may be declared irregularly, sometimes in response to personal cash needs rather than a formal capital policy. Documentation may be minimal, especially in earlier years.

From a corporate law standpoint, this informality is often manageable. From a tax standpoint, it creates uncertainty.

Section 55 analysis depends on what income has already been distributed and what income remains to support share value. Informal dividend practices make it difficult to reconstruct that history with precision. CRA audits frequently focus on whether prior dividends were paid out of income that the taxpayer now claims as safe, effectively double-counting the same income stream.

The problem is compounded when dividends are paid across related corporations without a clear articulation of purpose or source. In audit, CRA will often assume that prior distributions reduced safe income unless the taxpayer can demonstrate otherwise. Without contemporaneous records, that demonstration becomes difficult.

Informality, in this sense, is not a moral failing. It is a structural risk. Section 55 rewards documentation and punishes assumptions.

 

Emotional overlay on tax decisions

Family-owned enterprises are not purely economic actors. Decisions are influenced by relationships, expectations, and a desire to preserve harmony. These factors shape tax planning in ways that do not always align with technical optimality.

For example, dividends may be structured to achieve perceived fairness between siblings, to support retiring parents, or to address family disputes. Tax efficiency may be a secondary consideration. Advisors are often brought in late, after key decisions have been made.

Section 55 is indifferent to these motivations. A dividend paid for family reasons is analyzed the same way as a dividend paid for tax reasons. If the dividend reduces the inherent capital gain beyond what safe income supports, subsection 55(2) applies.

This disconnect between human decision-making and statutory analysis is a recurring theme in family enterprise disputes. CRA audits often read like morality plays, but the outcome turns on arithmetic and statutory interpretation, not fairness or intent.

For advisors, this means that family context must be acknowledged, but not allowed to obscure technical risk. The earlier section 55 is brought into the conversation, the easier it is to design solutions that respect both family dynamics and tax law.

 

Inter-generational transactions

Inter-generational transfers are a defining feature of family-owned enterprises and a prime breeding ground for section 55 exposure. These transactions often involve estate freezes, share reorganizations, internal buyouts, and holding company structures designed to facilitate succession.

Each of these steps can introduce dividends, redemptions, or deemed dividends that engage section 55. The risk is not confined to the transfer itself. It persists throughout the transition period, which may span many years.

One common issue is the misalignment between income and share value. After an estate freeze, income earned by the business may accrue economically to the new growth shares, while dividends continue to be paid on frozen shares. Without careful analysis, those dividends may be justified using corporate-level income that does not actually contribute to the capital gain on the frozen shares.

Another issue is timing. Inter-generational transactions are often staged. Safe income determination times may differ across steps, and income earned between steps may or may not be relevant. CRA audits in this area often focus on whether the taxpayer has improperly pooled income across periods or share classes.

Because these transactions are designed to preserve the family legacy, there is often a reluctance to revisit or unwind them. That reluctance increases the cost of a section 55 reassessment and limits remedial options.

 

Sale to a third party

The sale of a family-owned enterprise to a third party is often the moment when section 55 risk crystallizes. Years of accumulated income, informal practices, and inter-generational planning suddenly converge in a single transaction.

Pre-sale surplus extraction is common in this context. Owners want to remove excess cash, simplify the balance sheet, or optimize tax outcomes before selling shares. Dividends are paid. Structures are cleaned up. The focus is on closing the deal.

CRA’s focus, however, is retrospective. In audit, CRA examines whether the dividends paid in anticipation of the sale reduced the capital gain on the shares that were sold. If so, the safe income analysis becomes central.

What makes third-party sales particularly painful is that the transaction is final. The purchase agreement is signed. The proceeds are distributed. The business is gone. If CRA later recharacterizes a dividend under subsection 55(2), the tax cost falls entirely on the former owners, with no ability to renegotiate the deal.

For family-owned enterprises, this is often the first time section 55 is encountered in a concrete way. The reassessment feels unexpected because the planning felt conservative. In reality, the exposure had been building for years.

 

Conclusion to this section

Family-owned enterprises are uniquely exposed to section 55 not because they are aggressive, but because they are human. They grow organically, make decisions informally, and prioritize continuity over documentation. Section 55, by contrast, is mechanical, retrospective, and unforgiving.

For advisors, this creates both risk and responsibility. The risk is obvious: late-stage section 55 disputes are costly and disruptive. The responsibility is to recognize where that risk accumulates and to address it proactively, before transactions lock in outcomes.

Understanding why family-owned enterprises are exposed is the first step. The final section will translate that understanding into practical guidance for lawyers and advisors: when section 55 must be addressed explicitly, when silence is dangerous, and how to move from compliance to defensibility.

 

 

1. Key Takeaways for Lawyers and Advisors

Section 55 is not a provision that rewards familiarity. It rewards precision, timing, and restraint. For lawyers and advisors working with private corporations—particularly family-owned enterprises—the most dangerous section 55 files are not the exotic ones. They are the ordinary ones: internal reorganizations, dividends that “have always been done this way,” and transactions that feel commercial rather than tax-motivated.

This final technical section distills the practical lessons that emerge from the statute, the case law, and CRA’s current audit posture. It is written for professionals who need to know when to stop, when to escalate, and when historical comfort is no longer a defence.

 

When section 55 must be addressed explicitly

The first and most important takeaway is this: section 55 must be addressed explicitly whenever a dividend—actual or deemed—has the effect of reducing an accrued capital gain.

That statement sounds obvious, but in practice it is often misunderstood. Many files proceed on the assumption that section 55 is only relevant where the planning is aggressive, artificial, or designed to strip surplus. That assumption is wrong. Section 55 is triggered by effect, not by intent or sophistication.

In practice, section 55 should be addressed explicitly in at least the following circumstances:

When a dividend is paid between related corporations and the recipient corporation holds shares with an inherent capital gain. The deductibility of the dividend under subsection 112(1) does not resolve the issue. If the dividend reduces the value of the shares, section 55 is engaged.

When shares are redeemed or acquired for cancellation under subsection 84(3). Deemed dividends are not a workaround. They are a primary entry point into section 55 analysis and are frequently scrutinized by CRA.

When dividends are paid as part of a transaction, event, or series of transactions. This includes internal buyouts, estate freezes followed by redemptions, pipelines, pre-sale surplus extraction, and post-mortem planning. Series analysis matters, and silence on section 55 in these contexts is rarely defensible.

When dividends are paid shortly before a share sale or reorganization. Timing matters. CRA will ask whether the dividend reduced the inherent capital gain immediately before the dividend, and whether the income relied upon can reasonably be considered to contribute to that gain.

The common thread is this: if the dividend changes the capital gain picture, section 55 is no longer optional. It must be analyzed, documented, and supported.

 

When safe income requires independent review

Not every dividend requires a bespoke safe income study. But many more do than practitioners are comfortable admitting.

Safe income requires independent review when the answer is not obvious from first principles. If the conclusion relies on assumptions, proxies, or “rules of thumb,” it is already at risk.

Independent review is particularly important where:

The holding period is long and the income history is complex. Multi-decade ownership, multiple reorganizations, or incomplete records increase the likelihood that income has been double-counted, dissipated, or misattributed.

There have been informal or irregular dividend practices. Where dividends were paid without a clear source, timing discipline, or documentation, a fresh safe income reconstruction is often necessary to avoid retroactive assumptions.

There are multiple share classes or frozen shares. Share-specific analysis is not optional in these cases. Income that supports one class does not automatically support another.

The transaction involves a series of steps over time. Pipelines, staged reorganizations, and post-mortem planning are especially sensitive to safe-income determination time issues.

The dividend amount is material relative to the accrued gain. The larger the dividend relative to the gain, the more likely CRA is to scrutinize the contribution narrative.

Independent review does not mean over-engineering. It means forcing the computation and the explanation to stand on their own merits. In many cases, the most valuable outcome of an independent review is not a higher safe income number, but a clearer understanding of what portion of the dividend is actually defensible.

 

Why historical practice is dangerous

One of the most common—and most costly—errors in section 55 planning is reliance on historical practice. “We’ve always done it this way” is not a defence. In fact, it is often a red flag.

Historical practice is dangerous for three reasons.

First, the law has evolved. The post-2015 redesign of section 55 shifted the analysis toward results and contribution to capital gain. Practices that were tolerated, or simply not audited, in earlier periods may no longer survive scrutiny.

Second, CRA’s audit posture has evolved. CRA now frames its analysis explicitly around paragraph 55(2.1)(c). Even where the statutory language has not changed, the questions CRA asks have. Computations that do not address contribution to gain head-on are increasingly vulnerable.

Third, historical spreadsheets age badly. Safe income schedules are often copied forward year after year, accumulating adjustments and assumptions that no one remembers making. By the time a transaction occurs, the schedule may bear little resemblance to the statutory construct it purports to measure.

The danger is not that historical practice was malicious. It is that it was context-dependent. What worked in a low-audit environment, with less emphasis on share-specific analysis and contribution narratives, may fail under modern scrutiny.

For advisors, the implication is uncomfortable but necessary: comfort with precedent must yield to current defensibility. Every significant dividend should be assessed on its own facts, under current law and audit realities.

 

When to involve specialized tax advisors

Section 55 is one of those areas where early involvement of specialized tax advisors saves far more cost than it adds. The challenge is recognizing when generalist competence is no longer sufficient.

Specialized tax advisors should be involved when:

The transaction involves multiple steps, entities, or time periods. Section 55 analysis compounds quickly as complexity increases. What looks manageable at the outset can become unworkable without a coherent framework.

The dividend amount is material or irreversible. If the dividend cannot be undone, redeemed, or restructured after the fact, the margin for error is effectively zero.

The planning interacts with other high-stakes regimes. LCGE planning, post-mortem pipelines, surplus stripping concerns, and cross-border structures all raise the cost of a section 55 failure.

There is disagreement among advisors. If the corporate lawyer, accountant, and tax preparer do not align on the safe income position, that is a signal that the issue requires deeper analysis, not compromise.

CRA scrutiny is likely. Certain files—large dividends, pre-sale extractions, family reorganizations—are simply more likely to be audited. In those cases, the standard should be audit-defensibility, not filing-time acceptability.

Importantly, involving specialized tax advisors is not an admission that something is wrong. It is an acknowledgement that section 55 is a technical, evidentiary, and narrative problem, not just a computational one.

 

The professional risk perspective

For lawyers and senior advisors, section 55 is also a professional risk issue. It sits at the intersection of tax law, corporate transactions, and client expectations. When it goes wrong, the consequences are not limited to tax.

Clients often perceive section 55 reassessments as unexpected and unfair. The transaction “worked.” The tax was paid. The reassessment arrives years later. Advisors are asked why the issue was not identified earlier.

From a professional standpoint, the most defensible position is not that section 55 could not have been anticipated, but that it was anticipated and addressed. Even where the conclusion is that a dividend is safe, the existence of a documented analysis materially changes the risk profile.

This is why section 55 should be treated as a front-end planning issue, not a back-end compliance issue. Addressing it early protects the client, but it also protects the advisory team.

 

Conclusion to this section

For lawyers and advisors, the message of section 55 in 2026 is clear. The provision is not obscure. It is not optional. And it is not satisfied by historical comfort or administrative shorthand.

Section 55 must be addressed explicitly when dividends affect capital gain. Safe income requires independent review when the answer is not self-evident. Historical practice is dangerous because it drifts away from current law and audit reality. And specialized tax advisors should be involved not when things go wrong, but when the cost of being wrong is unacceptable.

Taken together, these principles shift the mindset from compliance to defensibility. That shift is not academic. It is the difference between a dividend that disappears quietly into history and one that resurfaces years later as a reassessment with no practical remedy.

The final conclusion will draw this together and return to the central theme of this article: safe income is no longer a spreadsheet exercise. It is a legal, economic, and evidentiary question—and it demands to be treated as such.

 

XI  Advanced Issues That Drive Outcomes in Modern Section 55 Audits

The analysis to this point establishes the doctrinal and computational foundation for a defensible safe income position. What remains—and what often determines the outcome in live CRA audits—are a small number of recurring “pressure points” that are underdeveloped in most professional literature but are repeatedly emphasized by CRA in audits, roundtables, and technical presentations.

This section addresses those pressure points directly. It does so not by introducing new doctrine, but by applying the existing statutory framework—particularly paragraph 55(2.1)(c)—to the fact patterns CRA actually challenges.

1. Phantom Income Revisited — Why the Debate Did Not End with Kruco

The term phantom income is often used imprecisely. In modern section 55 disputes, the real issue is not whether income was “real” or “unreal,” but whether it can reasonably be considered to contribute to the accrued capital gain at the relevant time.

Post-Kruco, two principles coexist:

1. Statutory safe income is tax income (subject only to subsection 55(5)); and
2. Contribution to gain is a factual inquiry, not a bookkeeping one.

Problems arise when these principles are collapsed.

CRA’s current audit posture does not generally argue that phantom income should be removed from statutory safe income. Instead, CRA argues that certain amounts—while included in tax income—cannot reasonably be considered to contribute to the capital gain, and therefore cannot protect a dividend under paragraph 55(2.1)(c).

Common examples include:

• Income offset by contemporaneous value destruction;
• Income generated by transactions that do not increase share value (e.g., circular flows);
• Income that is immediately extracted or neutralized by related outflows.

The defensible response is not to relitigate Kruco at Step 2. It is to accept tax income at Step 2, and then demonstrate—at Step 4—that the income relied upon actually explains the gain that existed immediately before the dividend.

This distinction is critical. Files are lost when practitioners argue phantom income at the wrong analytical layer.

1. Refundable Tax and Timing — Why “It Comes Back” Is Not Enough

Refundable dividend tax on hand (RDTOH) and Part IV tax frequently surface in section 55 audits, often implicitly rather than explicitly.

The recurring misconception is that refundable tax should not reduce safe income because it is, economically, temporary.

CRA’s position—consistent with its contribution-based framing—is more nuanced:

• At the safe-income determination time, tax paid is a real cash outflow.
• Refundability is contingent on future dividends, which may or may not occur.
• The statute asks what contributes to capital gain immediately before the dividend, not what might be restored later.

A defensible analysis therefore treats refundable tax as follows:

• Reduce safe income on hand for tax paid or payable at the relevant time;
• Do not “gross up” safe income for hypothetical future refunds;
• If refunds are later received before the dividend, they may restore contribution capacity—but only if they exist at the determination time.

This approach aligns with both CRA audit logic and paragraph 55(2.1)(c)’s timing requirement. It also avoids speculative adjustments that are difficult to defend under cross-examination.

1. DSI, Reorganizations, and ACB Misalignment — The Silent Failure Mode

One of the most common reasons section 55 files fail—even with technically correct safe income computations—is misalignment between safe income, share entitlements, and adjusted cost base following reorganizations.

This arises most often in:

• Estate freezes (s.86),
• Rollovers (s.85),
• Share exchanges (s.51),
• Amalgamations (s.87),
• Post-freeze dividend streams.

The failure mode is predictable:

• Safe income is computed at the corporate level;
• Dividends are paid on shares whose capital gain profile no longer tracks that income;
• CRA challenges contribution on a share-specific basis.

The correct analytical discipline is:

• Safe income must be allocated to shares based on economic entitlement, not historical ownership;
• ACB continuity must be mapped alongside safe income continuity;
• Where reorganizations sever the link between income and gain, safe income does not follow automatically.

This is not an academic refinement. It is one of CRA’s most effective audit arguments because it reframes the issue as a mismatch between income and value—precisely what section 55 is designed to police.

1. CRA-Style Examples (Applied, Not Hypothetical)

CRA consistently explains its section 55 positions through examples. A defensible practitioner resource should do the same. The following examples mirror CRA’s analytical style while remaining grounded in statute and jurisprudence.

 

Example 1 — Phantom Income vs Contribution to Gain

Facts
Opco earns $1,000,000 of taxable income over several years. During the same period, it incurs non-capital losses and writes off a failed expansion, destroying $900,000 of value. Immediately before a dividend, the accrued capital gain on the shares is only $100,000.

CRA Position
Only $100,000 of income can reasonably be considered to contribute to the capital gain. A $1,000,000 dividend triggers subsection 55(2) for the excess.

Defensible Analysis
Statutory safe income is $1,000,000. Safe income on hand—measured by contribution to gain—is capped at $100,000. The excess dividend is vulnerable.

 

Example 2 — Refundable Tax Timing

Facts
Holdco receives an intercorporate dividend and pays Part IV tax of $383,333. No refund has been triggered as of the dividend determination time.

CRA Position
The tax paid reduces the portion of income that contributes to share value at that time.

Defensible Analysis
Reduce safe income on hand by the tax paid. Do not assume refundability unless the refund exists at the determination time.

 

Example 3 — Estate Freeze Misattribution

Facts
After a freeze, Opco earns substantial income. Dividends are paid on fixed-value preferred shares issued on the freeze.

CRA Position
Post-freeze income accrues to growth shares, not frozen shares. Dividends on preferred shares reduce gain unsupported by income.

Defensible Analysis
Safe income must be allocated to shares that economically bear the income. Dividends on frozen shares may exceed their safe income.

 

Example 4 — Reorganization and ACB Drift

Facts
Shares are exchanged under s.85. Safe income is tracked corporately, but ACB resets on exchanged shares.

CRA Position
Income cannot protect dividends where the ACB/gain relationship has been severed.

Defensible Analysis
Reconcile safe income to post-reorganization ACB and accrued gain. If income does not explain the gain, it cannot shield the dividend.

 

Example 5 — Pre-Sale Dividend Timing

Facts
Dividend paid shortly before arm’s-length sale. Purchase price reflects cash on hand.

CRA Position
Dividend reduced inherent gain reflected in sale price.

Defensible Analysis
If income contributed to gain immediately before the dividend and was not already priced into the shares, it may protect the dividend. Otherwise, section 55 applies.

 

1. Why This Section Completes the “Coveted Research Tool”

With this section, the article now does what most professional commentary does not:

• It separates law from economics without collapsing them;
• It integrates post-2015 contribution framing explicitly;
• It addresses refundable tax timing honestly;
• It forces share-specific and ACB-aligned analysis in reorganizations; and
• It models CRA’s example-driven reasoning, rather than merely critiquing it.

This is precisely what allows the work to function as a planning tool, a review framework, and an audit-defence resource—not merely an explanation of section 55, but a guide to surviving it.

 

1. Conclusion — From Compliance to Defensibility

For many years, safe income was treated as a mechanical exercise—something to be “backed into” with a spreadsheet shortly before a dividend was declared or a reorganization closed. That era is over.

In today’s environment, safe income is no longer an accounting proxy or an administrative shortcut. It is a legal, economic, and evidentiary question that sits at the centre of section 55 risk. The statute demands more than arithmetic. CRA’s audit posture demands more than tradition. And the courts have made it clear that outcomes turn on whether the income relied upon can reasonably be considered to contribute to the accrued capital gain at the precise moment the dividend is paid.

What this means in practice is straightforward, even if the execution is not. Safe income must be computed with statutory discipline. It must be filtered for real economic survivability. And it must be reconciled—clearly and credibly—to the capital gain it is said to protect. Files that meet that standard tend to withstand scrutiny. Files that do not often fail long after transactions have closed, when remedies are no longer available.

For family-owned enterprises, the stakes are especially high. Long holding periods, informal dividend histories, internal reorganizations, and emotionally charged succession decisions create precisely the conditions in which section 55 issues surface late and expensively. For lawyers and advisors, the risk is not simply technical—it is reputational. Section 55 problems rarely announce themselves early, but they are unforgiving once they do.

This article was written to move the discussion from compliance to defensibility. Not to replace judgment, but to structure it. Not to promise certainty, but to provide a framework that aligns law, economics, and evidence in a way that can be explained—to CRA, to a court, and to your client.

At Shajani CPA, this is how we approach section 55 planning, safe income reviews, and audit defence. We work alongside family business owners, corporate and commercial lawyers, and professional advisors to ensure that surplus extraction, reorganizations, succession planning, and liquidity events are not only technically correct, but resilient under scrutiny.

Tell us your ambitions—succession, liquidity, growth—and we will guide you there, safely.

 

This information is for discussion purposes only and should not be considered professional advice. There is no guarantee or warrant of information on this site and it should be noted that rules and laws change regularly. You should consult a professional before considering implementing or taking any action based on information on this site. Call our team for a consultation before taking any action. ©2026 Shajani CPA.

Shajani CPA is a CPA Calgary, Edmonton and Red Deer firm and provides Accountant, Bookkeeping, Tax Advice and Tax Planning service.