Safe withdrawal rates and stock-market valuation
The famous 4% rule applies the same withdrawal whether you retire cheap or dear. At today’s US CAPE of 41.4, a 30-year history points to a sustainable rate nearer 4.5% on average — and as low as 2.9% to survive the worst past sequences. Here is why, and what it does to the numbers.
By Dominic Roe · Data Engineer & Business Intelligence Developer
Valuation-aware safe withdrawal rate (typical 30-year history)
4.5% / yr
≈ £22,588 a year (£1,882/mo), in today’s money.
Cautious end
2.9%
£14,569/yr · near the worst past cases
The flat 4% rule
4.0%
£20,000/yr · ignores valuation
Every 30-year retirement since 1881
Each dot is one starting month: its CAPE then, and the highest constant real withdrawal a 100%-equity pot could have sustained for 30 years.
Rule: safe rate = 3.25% + 0.53 × (1/CAPE). R² = 0.48 over 1365 starting months (1881–1994). The dashed line marks 4%. Today's CAPE as of 2026-06.
The 4% rule, and what it leaves out
The 4% rule is a fine starting point: withdraw 4% of your pot in the first year, rise with inflation thereafter, and history says a balanced portfolio lasts the distance. But it was calibrated across all starting points at once — cheap markets and expensive ones averaged together. The price you happen to retire at is missing from the rule, and that price matters, because of sequence-of-returns risk: a weak first decade does far more damage when you are selling units to live on than the same decade would later. Retire into an expensive market and you face lower expected returns and a higher chance of that bad-first-decade sequence at the same time.
What history actually supported, by starting valuation
For every starting month back to 1881 we worked out the highest constant, inflation-adjusted withdrawal a 100%-equity pot could have sustained for 30 years — then grouped those by how expensive the market was at the start. The pattern is stark.
| Starting CAPE | Typical safe rate | Worst case | Starting months |
|---|---|---|---|
| Under 10 | 9.6% | 6.0% | 229 |
| 10–15 | 7.8% | 4.5% | 458 |
| 15–20 | 6.4% | 4.1% | 497 |
| 20–25 | 4.7% | 3.7% | 167 |
| 25–30 | 3.6% | 3.3% | 12 |
| 30+ | 3.2% | 3.1% | 2 |
Cheap starts supported withdrawals around 8–10%; the most expensive starts, closer to 4% even in the typical case and into the low 3s at worst. Note the right-hand column: the market was rarely this expensive in the past, so the high-CAPE rows rest on very few retirements. Today’s valuation sits at the thin top edge of this record — the calculator’s line is partly extrapolating beyond it.
This is the retirement-side mirror of what valuation does to expected returns — and another angle on what a high CAPE does and doesn’t tell you. You can see where today sits across several gauges on the US valuation dashboard.
The caveats — read these before you act on it
- Equities only. This is a 100%-stock backtest, to isolate valuation. Real retirees hold bonds, which has historically softened sequence risk and often supported a slightly higher safe rate — so the figures here are the equity core, not a full plan.
- Sparse data at high valuations. CAPE was rarely above 25 in the windows with a full 30 years behind them, so the estimate at today’s level leans on the fitted rule more than on lived history.
- Constant real spending is rigid. The backtest never cuts spending. Retirees who trim withdrawals after bad years (a flexible or “guardrails” approach) have historically sustained higher starting rates.
- Costs and taxes are excluded, as is any state or workplace pension income that would reduce what your pot must cover.
- The past is not a guarantee. These are historical outcomes for US equities, presented as information and context — not a forecast and not financial advice. See the methodology.
Frequently asked questions
What is the 4% rule?
It is the best-known retirement rule of thumb: withdraw 4% of your pot in year one, then increase that pound amount with inflation each year, and a balanced portfolio has historically lasted at least 30 years. It comes from work by William Bengen and the Trinity study. Its blind spot is that it applies the same 4% whether you retire into a cheap market or an expensive one.
Does a high CAPE really mean I should withdraw less?
Historically, yes. Retiring into an expensive market combines lower expected returns with sequence-of-returns risk — a poor first decade does outsized damage when you are also drawing down. In our backtest the highest sustainable 30-year withdrawal was strongly related to the starting valuation, with the cyclically adjusted earnings yield (1/CAPE) explaining roughly half the variation. The cheapest starting points supported rates near 10%; the most expensive, closer to 4% or below.
Is this based on a 100%-stock portfolio?
Yes — it uses Robert Shiller's real total return for US equities only, so it isolates the effect of valuation cleanly. A real retiree usually holds some bonds, which has historically smoothed sequence risk and often supported a slightly higher safe rate than 100% equities alone. Treat the figures as the equity-driven core of the question, not a whole-portfolio plan.
How is the safe rate calculated here?
For every starting month in Shiller's history we compute the SAFEMAX: the highest constant, inflation-adjusted withdrawal a 30-year retirement could have sustained without running dry. We then fit that against the starting earnings yield to get the rule shown on the chart. The whole thing is regenerated from the source data by a script — no hand-typed figures — the same approach behind our expected-returns model.
So is 4% still safe today?
At today's elevated CAPE the cautious end of our estimate sits a little below 4% while the typical historical experience sits a little above it, so 4% is no longer the comfortable margin it was at lower valuations. Two honest caveats pull in opposite directions: valuations this high are near the edge of the historical sample, so the figure is partly an extrapolation; but adding bonds and flexible spending has historically helped. This is information, not advice.
Is the market cheap or expensive? Get the monthly read.
Once a month: the global CAPE and where it sits in history, the cheapest and most expensive markets, and what 10-year return today’s valuation implies. Free, no spam, unsubscribe anytime.
We use a double opt-in and never share your address. See our privacy notice.