What Is a Safe Withdrawal Rate?
Safe withdrawal rate is the percentage of a portfolio that can be taken out annually without exhausting it over a planned retirement. The figure depends far more on the order returns arrive in than on their average, because selling into a decline removes units that cannot participate in the recovery.
While you are adding money to a portfolio, the order returns arrive in does not matter — only the compounded total does. The moment you start taking money out, that stops being true, and it stops being true in a way that decides whether the money lasts.
How it works
A withdrawal rate is an annual percentage of the starting portfolio, usually raised each year with prices. Four percent of a million is forty thousand in year one, and then forty thousand in today’s money every year after.
The reason order matters is that selling is irreversible. Taking money out during a decline means selling more units to raise the same amount, and those units are gone — they cannot participate in the recovery that follows.
So a poor first decade is the scenario that ends portfolios, and it does so even when the average return over the full period is perfectly adequate.
The identical sequence reversed is survivable. Good years first build a balance large enough that the bad years arrive against a bigger base and a smaller proportional withdrawal. Same returns, same average, opposite outcome.
A worked example
Use the drawdown behaviour measured on this site’s shared series. Ninety-five percent of bars sat below a prior peak. The deepest drawdown ran 3.76%. The longest wait for a new high was 73 bars — in a stretch that finished up 3.61% overall.
Read that as a retirement rather than a trade. Being below a previous high is not the exception, it is the normal state — and a withdrawal plan that only works when the portfolio is at a high is a plan that works 5% of the time.
The long flat stretch is the real test. Seventy-three bars without a new high, while withdrawals continue every year, is the mechanism that turns a temporary decline into a permanent reduction in what the portfolio can support.
Nothing in that requires a crash. It requires an ordinary, measured, unremarkable period of going nowhere — which is a thing markets do routinely and plans rarely model.
Two different rules
A fixed real amount and a fixed percentage are not the same plan. Taking 4% of the starting balance and raising it with prices gives a predictable income and lets the portfolio run down in bad sequences. Taking 4% of the current balance every year can never exhaust the portfolio and produces an income that falls exactly when markets do.
The first prioritises the person, the second prioritises the pot. Most published figures describe the first, and most people assume they describe the second — which matters enormously, because only one of them can actually run out.
What comes out before you do
Every annual charge is deducted ahead of your withdrawal. On this site’s thirty-year fee measurement, 75 basis points costs 20.2% of the final balance and 150 costs 36.5%. A portfolio paying the higher figure supports a materially smaller withdrawal for the same risk of running out.
Rebalancing costs a round trip each time — 0.0098 on this site’s series, about 2% of the median bar range. Small per event, repeated annually for decades.
And the withdrawal itself has to grow. A fixed cash figure is a shrinking real income — at 3% inflation, purchasing power falls to about 41% over thirty years, which inflation covers in full. A plan that does not raise the payment annually is not a plan for the same standard of living.
The original data
On this site’s shared series: 95% of bars below a prior peak, deepest drawdown 3.76%, longest recovery 73 bars, over a stretch that finished +3.61%.
Thirty-year fee drag: 1.5% of the final balance at 5 basis points, 5.8% at 20, 20.2% at 75, 36.5% at 150.
Put the two together and the structure of the problem is visible. The portfolio spends almost all its life below a prior high, a fee is deducted every one of those years, and a withdrawal is taken on top — with the amount required rising annually with prices. A safe rate is whatever survives that combination, and it is lower than the number that survives an average.
When it fails
The characteristic failure is treating a published percentage as a rule rather than a result. A withdrawal figure is the output of a specific study, over a specific market history, at a specific asset allocation, with a specific assumption about fees and a specific definition of success. Applied to a different portfolio in a different period with a different cost structure, the number carries none of that and is simply a percentage somebody remembered. The arithmetic does not transfer just because the figure does.
A second failure is fixing the cash amount and never raising it, which is a plan for a steadily poorer retirement.
A third is ignoring fees when choosing the rate, when they are deducted before every withdrawal and compound for the whole period.
A fourth is assuming flexibility is available. A plan that quietly depends on reducing withdrawals in bad years should say so, because that is a different plan with a different failure mode.
And a fifth is running the numbers on averages. The average outcome is not the thing being planned for; the bad sequences are, and they are invisible to any calculation that uses a single return figure.
Related
Inflation covers why the withdrawal must rise every year to mean the same thing. Risk of ruin covers the same arithmetic of surviving a bad sequence. And asset allocation covers the mix that decides how deep the bad years go.
The arithmetic that surprised me is that two retirements with identical average returns can end completely differently, and the only difference is which decade the bad years landed in. While you are saving, order does not matter at all. The moment you start selling, it becomes the thing that decides the outcome.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.