How Much Is Enough: The 4% Rule and Its Critics Advanced

The most quoted number in retirement planning came from a study of American data in the 1990s, and every assumption it rests on is worth knowing before you build a life on it.

Where the 4% rule came from

The rule originates in work by William Bengen and, most famously, the Trinity study of the late 1990s. Researchers took historical US market returns, ran a retiree withdrawing a fixed percentage of their starting portfolio — adjusted each year for inflation — and asked how often the money lasted 30 years.

At a 4% initial withdrawal rate, with a portfolio of roughly 50–75% stocks, the money survived 30 years in about 95% of the historical periods tested.

Two things are frequently misunderstood. The 4% applies only to the first year; after that you increase the dollar amount by inflation, not by recalculating 4% of a changed balance. And "success" was defined as not running out within 30 years — in many of those periods the retiree died with several times their starting portfolio, because the rule is calibrated to survive the worst historical case rather than the typical one.

The assumptions behind it

Every one of these is worth checking against your own situation:

• US market data. The twentieth-century US was arguably the most successful equity market in history. Applying its returns to a global or Canadian portfolio is optimistic; several countries' investors would have fared far worse.
• A specific allocation. The result depends on holding a substantial equity weight throughout retirement. A conservative 30% equity portfolio does not support 4%.
• No fees. The study used index returns. A 1% advisory fee plus a 1% fund fee turns a 4% withdrawal into an effective 6% drain on the portfolio.
• No taxes. A Canadian withdrawing from a RRIF pays tax on every dollar. Withdrawing $40,000 of spending money may require withdrawing $50,000.
• A 30-year horizon. Retire at 55 and you may need 40 years, at which point 4% is meaningfully less safe.
• Rigid spending. Real retirees cut back in bad years. That flexibility is not in the model — and it is a large, unmodelled safety margin.

The 25× target, Canadian version

The most common error Canadians make with this rule is applying 25× to their entire desired income. You do not need your portfolio to fund the part CPP and OAS are already covering.

A useful cross-check: most retirees find they need 60–80% of pre-retirement income rather than 100%, because the mortgage is often gone, the commute has stopped, the children are independent, and — not trivially — you are no longer saving 20% of your income. That last point catches people: a household saving 25% was already living on 75%, so replacing 100% of gross income would fund a lifestyle upgrade rather than a continuation.

Sequence-of-returns risk — the thing that actually breaks plans

During accumulation, the order of returns barely matters — only the average does. During withdrawal, the order can matter more than the average, and this is the single largest risk to a retirement.

The mechanism: when you withdraw during a decline, you sell more units at low prices to raise the same dollars. Those units are permanently gone and cannot participate in the recovery. Early losses therefore compound into a permanently smaller portfolio in a way that identical later losses do not.

This is why "my portfolio averages 7% so I can withdraw 6%" is dangerous reasoning. Averages describe accumulation. Withdrawal is path-dependent.

Defending against it

• Hold one to two years of spending in cash so a market decline never forces a sale at the bottom. This is Lesson 7.1's emergency fund doing its most important job.
• Be flexible in the first decade. Skipping an inflation increase after a bad year is a small sacrifice with a disproportionate effect on survival.
• Reduce equity exposure approaching retirement, then raise it again — the “rising equity glidepath” — so the vulnerable years carry less risk.
• Cover essential spending with guaranteed income. Deferring CPP to 70 and considering a partial annuity (Lesson 8.3) removes sequence risk from the part of your budget that cannot flex.

Guardrails and variable spending

The modern refinement abandons the fixed inflation-adjusted withdrawal in favour of rules that respond to what the portfolio actually does. A common structure:

• Start at a withdrawal rate of roughly 4–5%.
• If the current withdrawal rises above an upper guardrail — say 20% above the initial rate, because the portfolio has fallen — cut spending by around 10%.
• If it falls below a lower guardrail because the portfolio has grown, increase spending.
• Otherwise adjust for inflation and change nothing.

Guardrails typically permit a higher starting withdrawal than 4% while improving the odds of not running out, because they respond to bad sequences instead of ignoring them. The trade is that your income varies, which requires enough discretionary spending in your budget to absorb a 10% cut without hardship.

Common questions

How much do I need to retire in Canada?

Take your desired annual spending, subtract what CPP and OAS will provide, and multiply the remainder by 25. For $60,000 a year with average CPP ($925.35/month) and full OAS ($742.31/month), government income is about $20,012, so the portfolio must produce roughly $40,000 — a target near $1.0 million. Adjust upward if most of your savings sit in a RRIF, since withdrawals are fully taxable, and downward if a large share is in a TFSA.

Is the 4% rule still valid?

It remains a useful benchmark for the right order of magnitude, but it was built on twentieth-century US data with no fees, no taxes and a rigid 30-year horizon. Its main weakness is sequence-of-returns risk: poor returns in the first decade can exhaust a portfolio that the same returns in a different order would have preserved. Most planners now use guardrails — adjusting spending when the withdrawal rate drifts too far from target — which allows a somewhat higher starting rate with better odds.

What is sequence-of-returns risk?

The risk that poor returns early in retirement do disproportionate damage, because withdrawing during a decline sells more units at low prices and those units never participate in the recovery. In a worked example with 25 identical annual returns and a 4.60% average, the portfolio that met its bad years first ran out in year 16, while the one that met them last still had $593,747 after 25 years. It is the reason average returns are a poor guide to withdrawal planning.