Three withdrawal rates, three different retirements
The 4% guideline came from historical US data on a 30-year horizon with a balanced portfolio. It is a reasonable starting point and a poor universal rule. Longer retirements, higher fees, weaker starting valuations and a desire for a larger safety margin all argue for something lower.
| Rate | Capital per $1,000/mo | Suits | Trade-off |
|---|---|---|---|
| 4.0% | $300,000 | Retirement at 65, ~30-year horizon | Requires flexibility in bad years |
| 3.5% | $343,000 | Early retirement, 35–40 years | Roughly 14% more capital needed |
| 3.0% | $400,000 | Very long horizons or low risk tolerance | A third more capital, or working longer |
The flexibility discountA retiree willing to cut discretionary spending by 10% in bad years can sustain a materially higher rate than one with a rigid budget. Adaptability is worth more than an extra percentage point of return.
Sequence risk: the thing this model cannot show
A constant-return projection hides the single largest danger in retirement. Two retirees with identical average returns can end up in completely different places depending on when the bad years arrive. Poor returns in the first five years of drawdown, while withdrawals are being taken, permanently impair the portfolio — the same losses arriving a decade later are largely survivable.
- Hold two to three years of spending in cash and short bonds so you never sell equities into a decline.
- De-risk on a glide path in the five years before retirement, not on the day itself. See allocation by age.
- Define a spending rule in advance — for example, no inflation increase in any year following a negative return.
- Delay guaranteed income where possible; deferring a state or social pension is often the cheapest longevity insurance available.
- Recheck annually, not daily. The plan should change with circumstances, not with headlines.
The healthcare gapRetiring before state healthcare eligibility can add a five-figure annual cost. Model it explicitly in your spending figure rather than assuming it away — it is the most common reason early retirement plans fail in year one.
How to use the result honestly
- Use the real-terms numberThe nominal projection is not comparable to today's spending figure. Compare like with like or you will overestimate by decades of inflation.
- Test the fee sliderHalf a percent of fees over thirty years frequently equals two extra working years. That is the actual price of an expensive platform.
- Run a pessimistic caseReturn at 5%, inflation at 3.5%. If the plan still works, it is robust. If it collapses, you have found the fragility now rather than at 67.
- Convert gaps into actionsA shortfall has four levers: save more, work longer, spend less, or accept more risk. Only the first three are reliable.
Limits, errors and borders of the standard advice
Critical reviewRetirement projections present a single confident number produced from assumptions that are individually uncertain and compound multiplicatively. The rules of thumb built on top of them are more fragile than their popularity suggests.
"The 4% rule tells you how much you can safely withdraw."
The claimWithdrawing four per cent of the initial balance, inflation-adjusted, is sustainable for a thirty-year retirement.
Where it breaksIt came from a study of one country's historical returns, over thirty-year periods, with a specific portfolio, ignoring fees and taxes — the most successful equity market of the twentieth century, which is a survivorship issue. It assumes rigid inflation-adjusted spending, which nobody actually does, and it says nothing about longer retirements or the valuation environment at the start. It is a research finding about historical worst cases that became a planning rule, and the rule now carries far more confidence than the finding supports.
The borderUseful as a rough sizing check on whether a target is plausible. Replace it in execution with flexible withdrawals — spending less after poor years and more after good ones — which historically supports a higher average rate with less risk of ruin than any fixed percentage.
"You'll need 70–80% of your pre-retirement income."
The claimA replacement-rate percentage of final salary estimates required retirement income.
Where it breaksSpending in retirement is not a percentage of prior income; it is a function of housing status, health, family obligations and how you intend to live. Someone retiring with a repaid mortgage and no dependants may need far less; someone facing rent, care costs or supporting adult children may need more than they earned. Real spending patterns also change shape over retirement — often higher early, lower in the middle, rising again with care needs — which no flat percentage captures. The rule persists because it lets a projection run without asking hard questions.
The borderA replacement rate is acceptable for a first estimate decades out. As retirement approaches, build the number from actual expected expenses, and separate essential spending — which should be covered by guaranteed income where possible — from discretionary spending, which can flex.
"Assume a 7% average return and project forward."
The claimLong-run historical averages provide a reasonable planning return.
Where it breaksAn average return applied uniformly is the single most misleading element of most projections, because the order of returns matters enormously once withdrawals begin — the same average with poor early years can exhaust a portfolio that the average alone suggests is safe. Historical averages are drawn from favourable markets and periods, are usually quoted before fees and often before inflation, and say nothing about starting valuations. A single deterministic line also creates false precision: small changes in the assumed rate move the projected balance by enormous amounts decades out.
The borderUse a range rather than a point estimate, run a deliberately pessimistic case, and treat any projection as a direction of travel to be revised annually rather than a forecast. If the plan only works at the optimistic rate, it is not a plan.
"Save 15% of your income and you'll be fine."
The claimA fixed savings rate reliably produces an adequate retirement.
Where it breaksThe figure assumes an uninterrupted career starting young, consistent returns, and retirement at a conventional age — three assumptions that fail routinely through career breaks, caring responsibilities, illness and redundancy. Starting age dominates the outcome far more than the percentage: the same rate beginning fifteen years later produces a dramatically worse result, so a single number cannot be right for everyone. It also ignores whether the money is invested sensibly and what fees are being paid, either of which can matter more than the extra percentage points.
The borderFifteen per cent is a reasonable target for someone starting in their twenties with a stable career. Late starters need substantially more, or a later retirement date, or lower spending — and the honest arithmetic is better faced early than discovered at sixty.
This tool is a deterministic model, which means it inherits the exact criticism we make above: it produces a single figure from assumptions that deserve a range, and it cannot model sequence risk. Treat its output as a sizing exercise, not a forecast. It also knows nothing about your tax position, state pension entitlements or healthcare costs, all of which are jurisdiction-specific and can dominate the result — and our advice to plan pessimistically means some readers will save more than they needed and consume less life while doing it.
Frequently asked questions
Should I include social security or a state pension?
Subtract expected guaranteed income from your monthly spending figure and model only the gap. That is the portion your portfolio must actually cover, and it often reduces the required capital dramatically.
Is the 4% rule still valid?
As a planning anchor, yes; as a guarantee, never. It was always a historical backtest, not a law. Treat it as the midpoint of a range and choose your position within that range based on horizon, flexibility and fee level.
What return should I assume?
Use something below your portfolio's long-run historical average. Assumptions that flatter the plan produce plans that fail. Between 5% and 7% nominal for a diversified portfolio is defensible; anything above 8% is optimism with a spreadsheet attached.
Does this account for taxes?
No. Withdrawals from tax-deferred accounts are generally taxable, which can reduce spendable income materially. Model the wrapper mix separately using the account comparison.
Next step
Readiness is one input. The other two are the account structure holding your money and the asset mix inside it. Work through 401(k) versus IRA for the wrapper decision, then glide paths by decade for the allocation — and if income is your goal rather than total return, the dividend playbook covers the alternative route.
Model notes. Deterministic constant-return projection with a static withdrawal rate. It does not run Monte Carlo simulation, model taxes, or account for sequence-of-returns risk. Educational use only.