Withdrawal strategies · Article 54

When the 4% rule can be too optimistic for early retirement in Canada

Understand why a longer retirement, early market losses, taxes, pension timing and uncovered expenses can make a simple 4% target too optimistic.

Published by North of Numbers · · 6 min read

Educational purposes only. Examples are fictional and in Canadian dollars unless stated otherwise.

Start here

What can make a 4% starting withdrawal too optimistic for an early Canadian retirement?

A long retirement adds more withdrawals and more years of uncertainty. Early losses can damage a portfolio even when later returns recover, while taxes and fees reduce money available for living costs. CPP or QPP and OAS may help later, but they do not automatically fund the early bridge, and reduced earnings can change pension estimates. A realistic analysis tests those conditions and identifies spending adjustments instead of treating 4% as a guarantee.

What you’ll learn

  • A 30-year research horizon cannot automatically cover a 50-year retirement.
  • The sequence of returns matters once spending removes investment capital.
  • Later pensions, potential work and public health coverage need explicit limits.

Early retirement lengthens the promise

A household retiring at 45 and planning through 95 is considering 50 years of withdrawals. Retiring at 65 with the same endpoint means 30 years. A method tested for the shorter period does not automatically cover the extra two decades.

The additional years create twenty extra payments, with risks beyond their direct cost. They also extend exposure to inflation, changing family needs, investment costs and periods without public-pension income. A plan can withstand an early downturn yet become strained much later.

Bengen himself distinguishes very long retirement horizons from the familiar shorter-horizon discussion. His commentary is not a Canadian guarantee, but it is a useful reminder that the creator of the original research does not treat horizon as irrelevant. Bengen's early-retirement discussion

A positive average does not describe the withdrawal sequence

When there are no cash flows, the order of a set of returns does not change their compounded product. Once money is withdrawn, the order affects how much capital remains to participate in later returns.

A poor early result can be particularly damaging because spending continues while the balance is depressed. A later recovery applies to fewer remaining dollars. This does not mean that every early loss causes failure; it explains why an average return alone cannot establish affordability.

It also explains why a retirement model should show annual balances. A smooth line that grows at an assumed average and subtracts spending conceals the very variation a withdrawal study is intended to examine.

Fictional CAD example: reverse two returns

Assume a $1 million portfolio, $40,000 withdrawn at the start of each year, no contributions and no other income. Express all amounts in constant initial-year dollars. The returns below are real returns after investment costs; personal taxes are ignored. Inflation is therefore not added again to withdrawals in this calculation.

Compare the same two annual returns in different orders: negative 20% and positive 20%. These are invented teaching paths, not a forecast or a representative market sequence.

CalculationLoss firstGain first
Opening portfolio$1,000,000$1,000,000
After first withdrawal$960,000$960,000
End of year one$768,000$1,152,000
After second withdrawal$728,000$1,112,000
End of year two$873,600$889,600

Without withdrawals, either return sequence would turn $1 million into $960,000. With the stated withdrawals, the gain-first path finishes $16,000 ahead. Both delivered identical percentage returns; they applied those returns to different amounts of remaining capital.

The next $40,000 withdrawal would represent about 4.58% of $873,600 and 4.50% of $889,600. The starting rate was 4% in both cases. The comparison isolates timing risk rather than estimating a lifelong safe rate.

Change the horizon, then change the return assumption

Consider a separate deterministic test with the same $1 million and $40,000 annual real withdrawals. If real returns after fees were exactly zero, the money would fund 25 full annual withdrawals and nothing beyond them. A 50-year plan would need investment growth, lower withdrawals, later outside income or additional capital.

Zero real growth is a stress assumption, not a prediction. Its value is that the arithmetic can be understood without a simulator: $1,000,000 ÷ $40,000 = 25. A 4% starting figure is not, by itself, the reason money lasts 30 or 50 years; the modeled investment path supplies part of the funding.

Now reduce annual spending in this zero-real-return exercise to $30,000. The same capital funds 33 full payments with $10,000 remaining for a partial next payment. That is a meaningful improvement, but it still does not fund 50 full years. Sensitivity analysis should reveal the remaining gap rather than stop after finding an improvement.

Public pensions do not arrive at the early-retirement date

A Canadian retirement plan can include later CPP or QPP and OAS income where eligibility and amounts support it. The early bridge must still be funded before the modeled payments begin. Treating all retirement years as though those pensions already exist understates initial portfolio demand.

Lower future earnings may also change pension projections. Service Canada explains that CPP depends on earnings, contributions and starting age, with provisions that affect how low-earning periods are treated. A projection built around continued work should not be carried unchanged into a plan that stops contributions much earlier. Service Canada: CPP amount factors

For example, an assumed $20,000 annual pension starting in year 21 cannot fund a $20,000 gap in year three. It may improve later sustainability, but it is not cash available during the intervening years. A phased model keeps those obligations visible.

Before-tax capital and after-tax living costs must reconcile

Suppose a household needs $40,000 after tax and assumes an effective 20% tax on the withdrawals used to fund it. The required gross withdrawal is $40,000 ÷ 0.80 = $50,000. That is 5% of a $1 million portfolio, even though the household spends only 4% of the portfolio's opening value.

This is fictional tax arithmetic, not an estimate of a Canadian household's actual rate. Ordinary RRSP withdrawals are taxable, and tax withheld when money is withdrawn is not necessarily the final liability. CRA: RRSP withdrawal taxes

Fees must also enter consistently. A model using returns already reduced by investment costs should not subtract those same costs twice. A model using gross returns cannot silently assume the portfolio is costless. Neither situation is resolved by treating a quoted rate as money available for groceries.

Health coverage and flexibility need realistic limits

Canadian public health insurance changes the retirement-cost picture, but it does not make every health-related expense free. Government guidance identifies services that are not universally covered by basic provincial or territorial plans, including dental care, prescriptions, glasses and ambulance services. Government of Canada: health coverage

A longer retirement may also involve home modifications, family assistance or care needs that were absent from the opening budget. These should be treated as scenarios with explicit costs and coverage assumptions, rather than presumed away.

The ability to earn money again is another uncertain resource. Skills, health, caregiving and labour-market conditions may change. Potential work can be shown as a contingency; calling it guaranteed future income would make the base plan stronger on paper without making a job available.

A stress test should identify what must change

Build an annual schedule with the expected withdrawal, pension income, tax estimate and remaining portfolio. Test an early loss, several high-inflation years, lower later pensions and a longer horizon separately before combining them.

For a learning exercise, explain why the two-year example's $16,000 difference cannot be repaired by saying both paths averaged zero percent. The answer is that withdrawals changed the dollars exposed to each return; arithmetic average returns also differ from compounded growth.

A sound educational conclusion is conditional: this spending schedule works under these assumptions, and these events require a change. A national label, a round percentage or a long-term average cannot provide the missing conditions.

Evidence behind the explanation

Sources & checks

Linked sources were consulted on October 28, 2026. The examples are original educational calculations. Source consultation and arithmetic checks do not constitute independent professional review.

  1. Does the 4.7% rule still apply to people who retire much earlier in life?William P. Bengen · Accessed 2026-10-28

    Author explicitly distinguishes very long horizons; no empirical rate from this short answer is imported.

  2. How much you could receiveService Canada · Accessed 2026-10-28

    Contribution history, low-earning years and retirement commencement affect CPP estimates.

  3. Tax rates on withdrawalsCanada Revenue Agency · Accessed 2026-10-28

    Ordinary RRSP taxes and withholding versus final liability.

  4. Learn about health care in CanadaGovernment of Canada · Accessed 2026-10-28

    Basic public coverage does not universally include all dental, drug and other health expenses.

When this article needs another review
  • Pension contribution or commencement rules change.
  • Relevant tax or health-coverage rules change.
  • New long-horizon research materially changes the evidence discussion.

Check the linked official guidance before applying a rule that may have changed.

About this educational article

Published by North of Numbers and created with AI assistance using the sources listed above. This article has not been reviewed by a qualified financial professional. Sources, assumptions and hypothetical examples are shown so readers can examine the reasoning.

This article explains general concepts; it does not assess individual suitability or provide personal financial, investment, tax, insurance, mortgage or legal advice.

Read our editorial standards · Report a correction