FIRE glossary
Withdrawal Rate vs Return
Your withdrawal rate is the percentage you draw from your portfolio each year; your expected return is the percentage that portfolio is assumed to grow by — they're two separate numbers, and treating them as interchangeable is one of the fastest ways to get a required pot wrong.
Two different numbers, easily confused
A withdrawal rate answers "how much can I safely take out each year?" An expected return answers "how much do I think my portfolio will grow by?" They sound related, but conflating them — assuming a portfolio that grows at 7% on average can support a 7% withdrawal rate — is a common and costly mistake. Ember keeps them structurally separate: the safe-withdrawal-rate formula (requiredPot = grossSpend / swr, in swr.ts) never reads your real-return or volatility assumptions at all. The SWR is a number you choose (resolveSwr), not one the engine derives from how your portfolio is assumed to grow.
Why an average return overstates what's safe
An average masks the path. A portfolio can return an average of 7% a year and still contain individual sequences — a sharp early loss, a slow first decade — that would exhaust a pot drawing at a rate close to that average. Once you're withdrawing every year, you're selling more of your portfolio (in unit terms) when it's down and less when it's up, which drags the safe rate below the average return in a way a single headline figure never shows. This is exactly what Ember's Monte Carlo simulator is built to surface: instead of applying one average return, it runs a chosen withdrawal rate against many independently sampled return sequences and reports what share of them leave the portfolio solvent through the horizon — a direct, honest answer to "does my rate survive realistic variation?" rather than "does it survive the average?"
Where the two numbers do meet
They aren't entirely unrelated — a drawdown-to-zero required pot (a different withdrawal basis to the SWR one) is built explicitly from an assumed real return, using a present-value-of-annuity formula, and its own "implied withdrawal rate" then invites the same comparison against the SWR figure. Ember also flags a custom SWR that looks unusually aggressive (above roughly 4.5%) or unusually conservative (below roughly 2.5%) in the calculation trace, precisely because a bare percentage, detached from any check against return variability, is easy to set too optimistically.
Across borders
The two numbers get more tangled once money crosses a border. A portfolio held in one currency but spent in another carries an extra layer of return uncertainty — currency movement — that a single "expected return" figure never captures, and that volatility interacts with sequence risk exactly the way market volatility does. Moving tax residency can also change the after-tax return you actually keep on the same pre-tax growth (see the safe-withdrawal-rate term on how Ember grosses spend up for tax), so a withdrawal rate that looked sound against one country's tax regime can be tighter — or looser — against another's. None of that is a reason to guess a smaller number; it's a reason to check the SWR against a stress test rather than against a headline average return.
Common questions
If my portfolio has averaged 7% a year, can I safely withdraw 7%?
No. An average return describes the middle of a wide range of possible outcomes, not a guarantee every year lands near it. A retirement that starts with a few bad years — even inside a portfolio that averages 7% over the long run — can deplete a pot at a much lower withdrawal rate than the average implies. That's sequence-of-returns risk, and it's the whole reason a safe withdrawal rate sits well below a typical expected-return figure.
Does Ember calculate my withdrawal rate from my expected-return assumption?
No — they're kept deliberately separate. Your SWR (resolveSwr) is an input you set directly (or Ember's default), and requiredPot = grossSpend / SWR never reads your real-return or volatility assumptions. Those assumptions feed a different calculation — the drawdown-to-zero present-value formula and the Monte Carlo simulator — so the two concepts can't silently blend into one number.
How do I actually check whether my chosen rate holds up?
Not by comparing it to a single average-return figure. Run the plan's Monte Carlo simulation, which samples many different return sequences (not one average) and reports what share of them keep the portfolio solvent through your horizon, or check it against a historical backtest of real market sequences.
Related terms
See withdrawal rate vs return in your numbers
The free calculator gives a rough estimate; the full planner models your actual accounts, pensions, residency moves and taxes — with the maths behind every figure shown.