Safe Withdrawal Rate

A safe withdrawal rate converts a portfolio into a paycheck. You take a percentage of the starting balance in year one, give yourself an inflation raise each year after, and ask what starting percentage would survive a full retirement even under poor conditions. The number is not a property of markets alone; it is a property of your situation. A 50-year-old planning for a 45-year retirement, a 75-year-old planning for 15, and a retiree who can cut spending 15% in a bad year all have genuinely different sustainable rates from the same portfolio.

The best-known benchmark is the roughly 4% figure from William Bengen's 1994 research and the 1998 Trinity study, both of which tested inflation-adjusted withdrawals against historical U.S. market returns. That work is worth knowing because it established the method still used today. But it described one case (a 30-year retirement, a stock-heavy portfolio, rigid spending, and no fees or taxes), and treating its output as a rule is how the concept most often gets misused.

What actually determines the number. Six inputs do most of the work, and they push in both directions. Time horizon is the largest: a 45-year early retirement supports a materially lower rate than a 25-year one, because there are more sequences in which things go wrong. Asset allocation matters non-obviously — very conservative portfolios have historically supported lower rates, not higher, because they lose the growth needed to outrun decades of inflation. Fees come straight off the top; an advisory fee plus fund expenses reduces the sustainable rate roughly one-for-one. Taxes depend on which accounts the money comes from, so two retirees with identical portfolios and identical withdrawal rates can have very different spendable income. Other guaranteed income (Social Security, a pension, an annuity) changes the job the portfolio has to do, and a portfolio covering only discretionary spending can safely run a higher rate because failure is survivable. And spending flexibility is the input retirees most control: a willingness to trim in bad markets raises the sustainable starting rate more than almost any portfolio change.

Why the rate is far below average returns. Sequence-of-returns risk. A retiree withdrawing from a portfolio that falls early sells shares at depressed prices to fund spending, locking in losses that later recoveries cannot fully repair. The sustainable rate is therefore set by the worst historical sequences, not the average ones, which is also why, in most historical periods, a conservative fixed rate left retirees dying with several times their starting wealth. That spread between the median outcome and the worst case is the entire argument for flexibility rather than precision.

The four families of method. Each holds something different constant. Fixed real spending keeps your standard of living steady and lets the portfolio absorb all the uncertainty, the approach the original research tested. Percentage-of-balance takes a set share of the current portfolio each year, so it can never be depleted, but income falls exactly when markets do. Guardrails start from a spending figure and adjust it when the withdrawal rate drifts outside preset bands, capturing most of the flexibility benefit while keeping income reasonably stable. Floor-and- upside covers essential spending with guaranteed income and spends flexibly from investments, which changes the question from "will I run out" to "how much discretionary spending do I have." A variant of the percentage-of-balance approach simply follows the IRS required minimum distribution divisors, which has the appealing property of adjusting for remaining life expectancy automatically.

What the research says now. Work since Bengen, including his own later studies using broader asset classes, has generally supported somewhat higher starting rates for retirees who will adjust spending, while long early retirements, high fees, and low starting yields argue in the other direction. There is no professional consensus on a single number, and a page that offered one would be misleading you.

The method is always the same three steps: establish what the portfolio must cover after other income, choose a starting rate appropriate to the horizon and flexibility, then decide in advance how spending will respond to markets.

A hypothetical example. Rosa retires at 65 with a $1,200,000 portfolio, 60% stocks and 40% bonds. Social Security covers a good share of her essential spending, so the portfolio's job is the remainder. Starting from a 4% sketch, she withdraws $48,000 in year one. Inflation runs 3%, so a rigid fixed-real approach would take $49,440 in year two regardless of what markets did, and would keep taking the inflation-adjusted amount even after a 20% drop. Rosa instead adopts guardrails. She still starts at $48,000, but agrees with her planner in advance that if her withdrawal rate climbs above about 5% of the current balance she will trim spending 10%, and if it falls below about 3% she will give herself a raise. The decision made in advance is the point: it converts an unbearable in-the-moment judgment during a crash into a rule she already accepted. Because she is willing to adjust, she can defensibly start higher than a retiree committed to never changing course. Figures are illustrative.

Pros

• Turns an abstract nest egg into a concrete annual spending figure, which makes retirement planning tractable.
• Grounded in worst-case market history rather than optimistic average returns, so it is conservative by construction.
• Simple enough to sanity-check any retirement plan in one line of arithmetic, and it underlies the useful 25x savings target.
• Framing spending as a rate makes the trade-offs visible — the effect of fees, of retiring earlier, or of agreeing to be flexible.

Cons

• It is often quoted as a single number when it is genuinely a range that depends on the individual, and the 4% figure in particular carries assumptions that rarely match a real retiree.
• Any rate derived from history is an estimate about an unprecedented future, not a guarantee.
• A fixed real rule underspends in most scenarios — the cost of protecting against the worst one is years of unnecessary frugality.
• It says nothing about which accounts to draw from, and the tax consequences of that sequencing move the net result meaningfully.
• Real spending is not smooth: it tends to be higher early in retirement and can spike late for health costs, which a constant rate ignores.