Retirement Planning in Canada

Retirement planning in Canada is less about a single decision than about how several sources fit together, and in what order they are drawn.

This section covers the general framework and the risks that determine outcomes. Specific situations, an incorporated professional, a business owner approaching a sale, a real estate investor, have their own pages, because each faces a different problem rather than the same problem at a different scale.

What tax deferral actually is, where it exists in Canada, and why deferred is not the same as tax-free, is set out on tax-deferred growth.

The leveraged arrangement that borrows against an assigned policy for retirement income, what it depends on and how it fails, is on insured retirement plan.

Why the picture differs for a Canadian physician, with the late start and the professional corporation, is on doctor retirement plan.

Retirement where the wealth sits in property, with its illiquidity and its tax at death, is on real estate investor retirement planning.

What Canadian retirement income is actually made of

Most people assemble income from more than one of the following, and the tax treatment of each differs.

Government benefits. The Canada Pension Plan or the Quebec Pension Plan, which reflect what was contributed over a working life, and Old Age Security, which does not. Both can be started earlier or later than the standard age, and the choice of timing changes the amount permanently. Old Age Security is also subject to a recovery at higher income levels, which makes the timing of other withdrawals a live question rather than an administrative one.

Registered plans. An RRSP, converted to a RRIF or an annuity by the end of the year you turn seventy-one, with a minimum withdrawal required each year thereafter whether or not you need the money. A TFSA, from which withdrawals are not income and do not affect income-tested benefits.

Employer plans. A defined benefit pension, which pays a stated amount, or a defined contribution plan, which pays what the account supports. The difference between the two is not the size of the number; it is who carries the risk.

Non-registered savings. Taxed as earned, with capital gains, dividends and interest treated differently from one another.

A business or a property. For many Canadians the largest single asset, and the one least like income until it is sold, which introduces timing and tax questions the others do not have.

Permanent life insurance, where it exists, which is discussed below with the qualifications it deserves.

The ordering question

The technical work in retirement planning is less about which sources you have than about the sequence in which you draw them.

Drawing from a RRIF first reduces the taxable balance that must eventually be withdrawn anyway. Drawing from a TFSA first preserves flexibility but wastes the sheltered growth. Drawing from a non-registered account first can be efficient where the gains are modest. Delaying government benefits raises them permanently. Each choice affects the others, and the interaction is where an accountant earns their fee.

One ordering rule sits above the rest and this site states it repeatedly. For most Canadian households, unused registered contribution room is the more efficient home for surplus money and should be used before a permanent insurance contract is considered for retirement purposes. A practice that skips past that is not giving you the full picture, and this one is paid by commission on insurance, which is exactly why it says so here rather than in a footnote.

The risks that decide outcomes

Accumulation is about growth. Decumulation is about survival, and the risks are different.

Sequence of returns. Once withdrawals begin, the order returns arrive in matters enormously. Two portfolios with identical average returns produce very different outcomes if one suffers poor years early, because each withdrawal during a decline locks the loss in permanently. This is the single most underappreciated risk in retirement, and it is why the years immediately before and after retiring receive disproportionate attention.

Longevity. Planning to an average life expectancy means planning to run out half the time. The risk is not dying early; it is living longer than the money.

Inflation. A fixed income loses purchasing power every year. Over a retirement measured in decades the erosion is substantial, and a plan built on nominal figures overstates what it delivers.

Health and care costs. Frequently the largest unplanned expense, arriving at the point when earning capacity is gone.

Tax and benefit interaction. Withdrawals can trigger the recovery of income-tested benefits, which behaves like an additional marginal rate. A plan optimised for investment return and ignoring this can be worse than a simpler one that accounts for it.

Where permanent insurance fits, honestly

Three functions, each specific, and none of them is a substitute for a retirement plan.

Capital not subject to contribution limits. Registered plans are capped. Where room is used and surplus continues, a permanent contract is one place that capital can sit and grow without annual taxation, provided it remains exempt under Regulation 306, Income Tax Regulations. That treatment is conditional, not automatic.

Access without the tax consequence of a registered withdrawal. Requesting an advance against a contract is a different transaction from withdrawing from a RRIF, and it does not add to income for the year in the same way. It has its own cost and its own tax treatment, set out on how a policy loan actually works.

A death benefit, which is the primary purpose. In a retirement context this matters more than it first appears, because it allows other assets to be spent rather than preserved. A household holding capital back so that something remains for the next generation is funding an inheritance out of its own retirement. Coverage can do that job instead.

The qualifications belong with the functions. Participating whole life insurance is an insurance product and it is not an investment. Judged against a market portfolio as a way to grow money it usually compares poorly. It requires durable surplus cash flow and a long horizon, and it suits fewer people than are shown it.

The government layer, and what it actually replaces

Two programmes carry part of every Canadian retirement, and both are smaller than people assume.

CPP or QPP is based on contributions made across a working life. It replaces a proportion of earnings up to a ceiling, which means higher earners see a smaller share of their income replaced. An owner paying only dividends contributes nothing and accrues nothing.

Old Age Security is based on residency rather than contributions, and it is subject to a recovery tax above an income threshold. A single large income year, such as a business sale or a property disposition, can trigger that recovery, which is a timing consideration rather than a reason to avoid the transaction.

The Guaranteed Income Supplement is income-tested and is not available to households with meaningful other income.

Deferring either payment raises it. Both CPP and OAS pay more if started later, and the decision turns on health, other income and whether the money is needed now. It is one of the few genuinely free optimisations available and it is frequently made by default rather than by decision.

Neither programme is designed to carry a household, and for anyone accustomed to a professional or business income they replace very little.

Drawing down, which is harder than accumulating

The phase that decides outcomes and receives a fraction of the attention.

The order of withdrawals matters. Registered money is fully taxable on withdrawal. Non-registered money is taxed on realised gains. A TFSA is not taxed at all. Drawing in the wrong order across twenty years costs more than most product decisions save.

Minimums are compulsory. A RRIF requires a withdrawal each year from the year after conversion, whether or not the money is needed, and it is income for purposes of the OAS recovery tax.

Income smoothing is where the value sits. Realising income in a low year rather than a high one, staying under a threshold that matters, and using different sources in different years.

The lump that ruins a year is the risk. A property sale, a business disposition, a large registered withdrawal. Each can be staged.

And none of this requires a product. It requires an accountant, a plan written down, and a household that reviews it annually.

What actually goes wrong in retirement

Five, in rough order of how often they arrive.

Outliving the money. The risk no portfolio removes and the reason annuities exist at all.

A poor sequence of returns early in drawing, which removes capital that never recovers.

Health costs that are not medical. Treatment is publicly funded; the surrounding costs are not. Home adaptation, travel, and a partner reducing their own hours to provide care.

Long-term care, which is the largest single unfunded item in most Canadian retirements and is rarely modelled at all.

Cognitive decline, which puts any arrangement requiring active management at risk in exactly the years it needs it. A plan that depends on being managed at eighty-five should say so before it is adopted.

A sixth belongs beside them. Supporting adult children, which is now common enough to be a planning assumption rather than a contingency, and which almost no projection includes.

Long-term care, the item nobody funds

Named separately because it is the largest unfunded exposure in most Canadian retirements and it is almost never in a projection.

Public coverage varies by province and covers less than people assume. Facility care carries a resident charge, home care is limited, and the difference between what is publicly available and what a family actually wants is paid privately.

The duration is the problem rather than the monthly cost. A short period is absorbable. Several years is not, and it arrives at the point in a retirement when the capital was supposed to be running down rather than being consumed faster.

It usually falls on one spouse first, and the cost is met from assets meant to support both.

Four ways households meet it, and each has a real cost. Self-funding, which requires capital held back and therefore not spent on retirement. Long-term care insurance, which is a distinct product with its own underwriting and its own definitions. A critical illness policy, which pays a lump sum on diagnosis and is not the same thing. Family care, which is unpriced and is paid in somebody's working hours.

Model it before assuming it will not happen. A plan that works only if nobody needs care is not a plan; it is an assumption with figures attached.

Where permanent insurance genuinely fits in retirement

Stated narrowly, because this is the section a practice like this one is most tempted to overstate.

Estate liquidity. A deemed disposition at death produces a tax bill payable before assets can conveniently be sold. Coverage sized to that liability provides cash at the moment it is owed, so heirs are not forced to sell property or shares in a poor market. This is the clearest use and it is a funding job, not a growth one.

Permission to spend. A household intending to leave something behind often underspends its own retirement to protect the inheritance. Where coverage secures the legacy, the retirement capital can be used for the retirement. The benefit is behavioural and it is real.

A dependant who will always need support. The need does not expire, so the coverage should not either.

A business obligation that survives the owner.

And where it does not fit. As a way to grow money, where it usually compares poorly. As a substitute for registered plans. As an income source arranged through borrowing, which is a leveraged strategy examined separately on insured retirement plan including how it fails.

Four uses, narrowly stated. A description that finds more than four is probably selling.

How much is enough, and why the usual answers are unhelpful

A percentage of final salary is the common rule and it assumes spending follows income, which it does not. Two households on the same salary can need very different amounts.

A multiple of savings is the other, and it says nothing about what those savings must do.

The useful method runs the other way: start from spending.

What does the household actually spend now? Three months of statements answers it, and most people are wrong about the figure by a material margin.

Which of that stops at retirement? Commuting, professional fees, mortgage payments if the mortgage ends, and saving itself, which is often the largest single line and disappears entirely.

Which of it starts? Travel early on, health costs later, and support for adult children at unpredictable intervals.

What is already guaranteed? CPP or QPP, OAS, any defined benefit pension. Subtract it.

The remainder is what capital must produce, and it is the only number in this subject worth building a plan around.

Then ask what happens if it lasts ten years longer than expected, because that is the risk the arithmetic above does not carry.

An owner's position differs enough to need its own treatment, which is on the business owner's retirement plan.

Retirement is not one date

The framing that causes most of the disappointment.

It is rarely a single stop. Reduced hours, consulting, a partial sale, a period of caregiving that ends employment earlier than intended.

The early years cost more than the middle ones. Travel and activity front-load spending, then it settles, then health costs raise it again late. A projection assuming level spending across thirty years has described nobody.

And the timing is frequently not chosen. Health, redundancy or a family obligation decides it for a substantial share of people, which is the argument for having the plan work at several possible dates rather than one.

Plan for a range. A plan that only works if you retire exactly at sixty-five, in good health, having sold the business at the expected price, is a plan with a single point of failure at each of those four points.

The order that holds for most households

Not advice, and an ordering defensible in most circumstances.

Know what you spend. Everything downstream depends on it.

Fill the TFSA. Tax-free on withdrawal, fully liquid, and it does not count as income against the OAS recovery tax in retirement, which makes it unusually useful late.

Use RRSP room where the marginal rate makes the deduction worthwhile.

Decide the CPP and OAS start dates deliberately, rather than by default.

Write down the drawdown order before it is needed, with an accountant.

Model long-term care rather than assuming it will not arrive.

Then consider insurance, sized to a liability the earlier steps have identified, for one of the four uses named above.

Six of those seven generate no commission. That is the most useful thing this page can tell a reader about how to weigh the rest of this site.

The conversation worth having with a spouse

Retirement is usually planned as one household and lived as two people with different information.

Both should know what exists. Accounts, pensions, policies, where the documents are, and who to contact. The commonest crisis after a death is administrative rather than financial: the money exists and nobody can find it.

Both should know the drawdown plan, because the survivor will be executing it alone.

Both should understand what changes at the first death. Pension survivor benefits are frequently reduced. CPP survivor benefits are capped. Two OAS payments become one. Household costs do not halve.

And both should have agreed what happens if one needs care, before anybody does. That decision made in advance is a plan. Made afterwards it is a crisis with a bill attached, taken by whichever spouse is still able to take it, usually alone and usually quickly.

The plan that survives being wrong

A retirement plan is a set of assumptions about forty years, and several will be wrong.

Assume you live longer than expected. The error is asymmetric: money left over is a small problem and running out is not.

Assume the timing is not yours. Health, redundancy or a family obligation decides it for a substantial share of people.

Assume returns arrive in an unhelpful order at some point during drawing.

Assume care will be needed by somebody.

A plan that works under all four is conservative and dull. A plan that works only if none occurs is a forecast wearing a plan's clothing.

The three numbers worth knowing before anything else

What you spend. Three months of statements answers it, and most households are wrong by a material margin.

What is already guaranteed. CPP or QPP, OAS, any defined benefit pension.

The gap between them. That figure is what capital has to produce, and it is the only number in this subject worth building a plan around.

Everything else on this site is downstream of those three, and none of them requires a product to establish.

The failure these pages are written against

A household arriving at retirement having planned the accumulation and not the drawing.

Thirty years of contributions, and no decision made about withdrawal order, timing of government benefits, care, or what happens at the first death. Accumulating is the easier half and receives most of the attention, and the harder half arrives whether or not it was planned for.

What this section is for

Establishing what a Canadian retirement is actually made of, so that any specific plan can be checked against it. Nothing here requires a product, and most of the work that decides outcomes generates no commission for anyone.

What this section does not own

The comparison against registered accounts. Whether a permanent contract is preferable to a TFSA or an RRSP for a particular person is a comparison against a non-insurance alternative, which is an argument rather than a description. Those live in objections and risks, where the heavier disclosure applies and where the case against is stated at the same length as the case for.

The mechanics of the contract. Cash value, dividends, the adjusted cost basis and the exempt test belong to policy basics, which is the reference layer for this whole site.

The corporate analysis. How corporate surplus is taxed while held, and how a death benefit is credited to the Capital Dividend Account, belongs with business owners.

The specific situations

Each of these faces a different problem, which is why each has its own page rather than a section here.

The incorporated professional, whose retirement savings and business are the same pool of money, and whose contribution room is affected by how they pay themselves.

The business owner approaching a sale, for whom the retirement plan and the succession plan are one exercise, and for whom the timing of a sale can matter more than any investment decision.

The real estate investor, holding assets that produce income but are not liquid, and facing a deemed disposition on property that has appreciated for decades.

The insured retirement plan, which is a specific arrangement with specific requirements and specific ways of going wrong, and which is frequently described with more confidence than it deserves.

What to establish first

Before any product conversation, four things.

What the government benefits will actually pay, and when. Service Canada and Retraite Québec both provide estimates. Most people are surprised in one direction or the other.

What registered room remains. The notice of assessment states it. If there is meaningful unused room, that is the conversation to have first.

What the minimum withdrawals will be. A RRIF requires a withdrawal each year regardless of need, and for some households the forced withdrawal, not the spending, is what drives the tax bill.

Whether an inheritance is an objective. If it is, say so, because it changes the whole plan. If it is not, that changes it too, and a great many people have never been asked.

Only after those four does a question about any product become answerable.