RRIFs and decumulation — turning savings into income

An RRSP has an expiry date. At the end of the calendar year you turn 71 it stops being a place to accumulate and has to become something that pays you — and the rules set a floor under how fast. That floor, applied to a balance you spent four decades building, is the mechanic that quietly turns a successful savings plan into a tax problem nobody mentioned on the way in.

Decumulation gets a fraction of the attention accumulation does, and it has more moving parts: three kinds of account taxed three different ways, two public pensions with their own timing decisions, a benefit clawback keyed to a single line of your return, and a mandatory withdrawal schedule that ignores all of it. Most of the leverage, though, sits in a handful of choices — and several of them are made years before the deadline arrives.

The deadline, and the three ways through it

The RRSP must be wound up by December 31 of the year you turn 71. Not your birthday — the end of that calendar year. Three exits exist, and they can be combined in any proportion:

• Convert to a RRIF. The default, and what most people do. The account keeps its investments and its tax shelter; nothing is sold and nothing is taxed at conversion. What changes is that a minimum has to come out every year from the following year onward.
• Buy an annuity. Hands the longevity risk to an insurer in exchange for a defined income for life. Predictable, and irreversible.
• Withdraw the balance. Fully taxable in one year, which for any meaningful balance means the top bracket. Almost never right as a whole-account choice, occasionally right for a small remainder.

Doing nothing is the fourth option and the worst: the plan is deregistered and the entire balance becomes income in that year. Converting is administrative — a form at your institution — but the deadline is not, and not every institution chases you for it. The RRSP's whole design is deduct now, pay tax later, as the RRSP guide sets out. This is the "later", and it arrives on a schedule you do not choose.

How the minimum actually works

A RRIF minimum is a percentage of the account's value on January 1, set by your age on that same date. The percentage is fixed by regulation, it rises every single year, and it rises faster than most people picture: gently through the seventies, more steeply through the eighties, flattening only in the mid-nineties, where it stays level for the rest of the plan's life. Below 71 the factor follows a plain formula — one divided by ninety minus your age — which is why converting early produces a small required withdrawal rather than a large one.

Two things follow from the shape of that curve.

First, the dollar minimum can keep climbing even while the balance falls, because the percentage grows faster than a drawn-down portfolio typically shrinks. A plan that comfortably funds the first decade of retirement may be pushing out considerably more income than you want by your mid-eighties — income taxed at whatever bracket the rest of your return puts you in.

Second, there is no maximum. The minimum is a floor and never a ceiling: you can always take more, and in the low-income years you usually should.

One mechanical detail catches people every spring: nothing is withheld from the minimum. It is paid gross and the tax is settled at filing. Withdrawals above the minimum are withheld at graduated rates, the same as an RRSP withdrawal. So the cautious retiree taking exactly the minimum is the one most likely to owe a lump in April.

The younger-spouse election

The minimum is calculated from an age, and that age does not have to be yours. When the RRIF is established you may elect to use the age of a younger spouse or common-law partner instead — permanently, for the life of the plan. The account still belongs to you and the withdrawals are still your income; only the percentage changes.

With a spouse several years younger the effect is a materially smaller required withdrawal every year, which is exactly what you want if the goal is to keep money sheltered and keep control of your own taxable income. The election is made once and cannot be revisited, so decide it deliberately rather than accepting whatever the conversion form defaults to.

Withdrawal order is where the money is

Once the RRIF exists, the annual question is not how much to take from it. It is which tap each dollar of spending comes out of. A retiree with a RRIF, a TFSA and a non-registered account holds three taps with three different tax treatments, and blending them on purpose is worth more than any fund decision available at that age.

Take someone at 72 who needs $60,000 to spend, receives $22,000 from CPP and OAS, and faces a RRIF minimum of $14,000. Assume, purely for illustration, a combined marginal rate of 25% on taxable income up to $57,000 and 37% above it.

Strategy A is what happens when the RRIF is treated as the retirement account and the TFSA as an emergency fund left untouched. It costs about $15,000 more tax in a single year, entirely because income was pushed into a bracket that never had to be reached.

Strategy B looks like the winner and is the usual overcorrection. It pays the least tax this year — but it leaves the RRIF nearly intact, so next year's minimum is a higher percentage of a larger balance, and whatever survives to your final return is taxed there in one lump.

Strategy C is the version worth building a plan around. It fills the low bracket with RRIF income on purpose and tops up from the TFSA. It pays more tax than B this year, and in doing so moves $21,000 more out of the RRIF at 25% instead of leaving it to face 37% under a later forced withdrawal — or the top rate on a final return.

Run your numbersRetirement decumulation→

Two thresholds sit on top of the bracket table and behave like extra brackets. The OAS recovery tax takes back fifteen cents of every dollar of net income above its threshold, stacking directly on your marginal rate. The Guaranteed Income Supplement is reduced far more sharply than that — at a rate above any tax bracket in the country — so for a lower-income retiree an extra RRIF dollar can cost more in lost benefits than it ever saved as a deduction. When you start each pension feeds straight into this, which is why CPP and OAS timing is really a decumulation decision rather than a retirement-date one.

Sequence of returns

Everything above concerns tax. The other risk of the drawdown years is arithmetic of a different kind: the order your returns arrive in starts to matter enormously the moment withdrawals begin.

While you are contributing, order is irrelevant — a given set of annual returns produces the same ending balance in any sequence. Add withdrawals and that stops being true. Money taken out during a decline never participates in the recovery, so a poor first few years permanently shrinks the base every later good year compounds on.

Bad years firstGood years first

Same six returns, opposite order, $24,000 withdrawn each year

	Bad years first	Good years first
Start	400	400
Yr 2	283	495
Yr 4	247	479
Yr 6	282	347

Both portfolios start at $400,000, withdraw $24,000 at the start of each year, and earn the same six annual returns — three of −10% and three of +18% — in opposite orders. Balances in thousands.

The two portfolios earn identical returns and identical averages. Only the order differs, and it costs roughly $65,000 over six years on a $400,000 starting balance. Stretch the horizon and the gap widens; a long enough bad start ends in depletion rather than in a smaller balance.

The defences are unglamorous and effective. Hold a year or two of spending in cash or short bonds so a decline never forces a sale at the wrong moment. Keep the withdrawal flexible enough to trim in bad years — an early spending cut is worth several years of portfolio life. And do not carry an accumulation-era asset mix into the first decade of withdrawals, where the whole risk is concentrated. This is precisely what a fixed 4% rule assumes away by quoting a single number.

The years before 71 matter more than the years after

The best decumulation decisions are made before the RRIF exists. The stretch between leaving work and the conversion deadline is usually the lowest-income period of an adult life — no employment income, and possibly no pension income yet if you deferred it — which makes it the cheapest window you will ever get for moving money out of an RRSP.

Drawing the RRSP down deliberately in those years, at rates far below what a forced withdrawal eventually attracts, is the RRSP meltdown strategy, and for most retirees it is the largest single lever available. It compounds with everything else here: a smaller RRSP at 71 means smaller minimums, less clawback exposure, and a smaller balance sitting on the final return.

Two smaller mechanics belong in the same window. Pension income splitting lets a couple move up to half of eligible pension income — which RRIF income becomes at 65 — to the lower-income spouse, flattening two returns into something nearer to one. And the pension income amount is a modest credit a small RRIF withdrawal at 65 can claim where an RRSP withdrawal cannot, one of the few genuine reasons to convert part of a plan early.

None of this rewards reasoning one year at a time. Run the sequence forward instead — the retirement projection shows what arrives at the deadline, and the RRSP calculator shows how much you are still adding to it. The schedule is fixed; what you do about it, mostly, is not.