How Much Can You Safely Withdraw From Retirement Savings Each Year?
You spent thirty years learning to put money in. Now you need a rule for taking it out, and the rule matters more than most of the decisions that got you here — because a withdrawal plan that’s 2% too aggressive doesn’t fail slowly. It fails all at once, at 82, with no way to fix it.
The good news is the arithmetic is simple. The complication is that the arithmetic isn’t the risk.
Every figure below is a worked example with stated assumptions.
Where 4% Comes From
The 4% rule says: withdraw 4% of your portfolio in year one, then increase that dollar amount with inflation each year afterwards.
On a $1,000,000 portfolio that’s $40,000 in year one. If inflation runs 2.5%, year two is $41,000, year three $42,025, and so on — regardless of what the portfolio did.
The logic is that if a balanced portfolio returns roughly 5% above inflation over the long run, spending 4% leaves a 1% margin. The rule survived historical periods including some genuinely awful ones, which is why it stuck.
Two things about it are frequently misstated. It was designed around a 30-year retirement, so a retirement starting at 55 asks more of it than one starting at 67. And it was never meant as a rule you follow mechanically off a cliff — it’s a starting rate.
How Long a Portfolio Lasts at Each Rate
Assume a constant 5% real return, withdrawals rising with inflation, starting balance $1,000,000:
| Withdrawal rate | Year 1 withdrawal | Portfolio lasts |
|---|---|---|
| 3% | $30,000 | Indefinitely |
| 4% | $40,000 | Indefinitely |
| 5% | $50,000 | Indefinitely (exactly break-even) |
| 6% | $60,000 | ~37 years |
| 7% | $70,000 | ~26 years |
| 8% | $80,000 | ~20 years |
| 10% | $100,000 | ~14 years |
The shape of that table is the thing to notice. Below the return rate, the portfolio is a perpetual machine. Above it, every extra percentage point costs you far more than the last — going from 6% to 7% removes eleven years, and 7% to 8% another six.
It also flatters reality badly, because it assumes returns arrive as a smooth 5% every single year. They don’t, and that’s not a small technicality.
Sequence of Returns: Why the Order Beats the Average
Two retirees, both starting with $1,000,000, both withdrawing $40,000 a year, both experiencing exactly the same set of returns — in different orders.
Retiree A gets the bad years first: −20%, −20%, then recovery.
| Balance | |
|---|---|
| Start | $1,000,000 |
| Withdraw $40,000 | $960,000 |
| −20% | $768,000 |
| Withdraw $40,000 | $728,000 |
| −20% | $582,400 |
Two years in, the portfolio is at $582,400 and the $40,000 withdrawal is now 6.9% of the balance — well past sustainable. Retiree A is in trouble, and every future recovery has to work on a much smaller base.
Retiree B gets the good year first: +20%, then the same two −20% years.
| Balance | |
|---|---|
| Start | $1,000,000 |
| Withdraw $40,000 | $960,000 |
| +20% | $1,152,000 |
| Withdraw $40,000 | $1,112,000 |
| −20% | $889,600 |
| Withdraw $40,000 | $849,600 |
| −20% | $679,680 |
After three years and three withdrawals, Retiree B is at $679,680 — having taken $120,000 out to A’s $80,000 and absorbed the same two crashes.
Same average return. Different order. Permanently different outcome. That’s sequence of returns risk, and it explains why the first five to ten years of retirement carry risk that the twentieth year simply doesn’t. In accumulation, a crash is a discount. In drawdown, it’s a forced sale.
Three Ways to Handle It
Guardrails. Don’t defend a fixed dollar amount. Set a band — say, cut the withdrawal by 10% if the rate rises above 5% of the current balance, and allow a 10% raise if it drops below 3%. Retiree A’s guardrail would have triggered in year two, cutting the withdrawal to $36,000 and materially changing the trajectory. Most retirees have more spending flexibility than a fixed-withdrawal model assumes, and the flexibility is worth using.
A cash buffer. Hold two to three years of withdrawals in cash or short-term bonds. It exists so that in a crash you can spend the buffer rather than sell equities at the bottom, and refill it in good years. It costs you some expected return in exchange for not being a forced seller — a trade most people are happy to make once they’ve watched a portfolio fall 30%.
Front-load the flexible spending. If travel is going to be $10,000 a year in your first retirement decade and $3,000 in your third, that’s $7,000 a year of genuinely discretionary spending sitting in the early, high-risk window. Knowing which parts of your budget are load-bearing and which aren’t is what makes a guardrail workable rather than theoretical.
The Withdrawals You Don’t Choose
Two constraints will override your plan regardless of what you’d prefer.
Required minimum distributions. RMDs begin at age 73 for most people retiring now, rising to 75 for later cohorts under SECURE 2.0. They’re calculated on your traditional balance and an IRS life-expectancy divisor, and they don’t care whether you need the money. A large traditional balance can force taxable withdrawals well above what you planned to spend. Roth IRAs are not subject to RMDs, which is a strong argument for building one during your working years — and for considering conversions in the low-income window between retiring and 73.
Your first RMD can be delayed to April 1 of the following year, but doing so means two RMDs in one tax year. That’s usually a worse outcome than taking the first one on time.
Taxes. A $40,000 withdrawal is not $40,000 of spending money if it comes from a traditional 401(k) — it’s ordinary income. The conventional order is taxable accounts first, then traditional, then Roth last, on the logic that the tax-free account should compound longest. It’s a good default and a bad rule to follow blindly: in the years between retiring and RMDs starting, deliberately realising some traditional income at a low rate can save considerably more than it costs.
What to Put in the Spreadsheet
Inputs: starting balance, target first-year withdrawal, real return assumption, inflation assumption, expected Social Security and the age it starts.
Outputs: withdrawal as a percentage of current balance each year, ending balance by year, and the year the portfolio hits zero — if it does.
Then three scenarios: returns arriving smoothly, returns front-loaded with losses, and returns back-loaded. If your plan only survives the first one, it isn’t a plan.
And one schedule: projected RMDs from 73 through 90, so you can see the years where forced withdrawals exceed what you intended to spend before those years arrive.
The number you’re actually looking for isn’t the safe withdrawal rate. It’s the size of the gap between what your portfolio must produce and what your guaranteed income already covers — which is where the whole calculation starts. That’s the complete retirement calculator build, and the Social Security claiming decision that sets the guaranteed side is here.
Featured on ReadySheetGo
Retirement & FIRE Calculator — 9 tabs, 463 working formulas. The Withdrawal Strategy tab handles exactly this: 4% rule analysis against your own balance, a Required Minimum Distribution schedule for ages 73 through 90, and a tax-efficient withdrawal order guide across account types. The Scenarios tab runs best, expected and worst case side by side so you can see what a bad first decade does before it happens, with a Social Security claiming comparison at 62, 65, 67 and 70. Also includes a Dashboard with ON TRACK / NEEDS ATTENTION status, a 40-year Savings Projection with chart, an Account Planner across 401(k), Traditional IRA, Roth IRA, HSA and brokerage, an inflation-adjusted Expense Planner by category, and a FIRE Calculator. Works in Excel and Google Sheets, no macros. Instant digital download — $17.99.
Frequently Asked Questions
Is the 4% rule still safe?
It's a reasonable planning anchor, not a guarantee. The arithmetic is sound — at a 5% real return, a 4% withdrawal is sustainable indefinitely because you're spending less than the portfolio produces. What it doesn't survive is a bad first decade. Withdrawing a fixed dollar amount from a portfolio that drops 20% two years running turns a 4% withdrawal into a 5.2% one, and the portfolio never fully recovers. Most people treat 4% as a starting rate to be adjusted, not a rate to be defended.
What is sequence of returns risk?
It's the fact that two portfolios with identical average returns can end up in completely different places depending on which years the losses land in. In the worked example here, a portfolio that loses 20% twice at the start falls to $768,000 while one that gains 20% first before the same losses is still above $1,000,000 — same withdrawals, same returns, different order. Losses early in retirement do permanent damage because you're selling assets to fund withdrawals at exactly the wrong price.
How long will my retirement savings last at a 5% or 6% withdrawal rate?
At a constant 5% real return, a 5% withdrawal rate lasts indefinitely because withdrawals exactly match returns. A 6% rate depletes the portfolio in about 37 years, 7% in about 26 years, 8% in about 20 years and 10% in about 14 years. Those are smooth-return figures and real markets aren't smooth, so treat them as an upper bound on how long you'd have rather than a forecast.
When do required minimum distributions start?
For most people retiring now, RMDs begin at age 73. Under SECURE 2.0 the applicable age rises to 75 for later cohorts. Your first RMD can be deferred to April 1 of the following year, but that means taking two in one tax year, which often pushes you into a higher bracket. RMDs apply to traditional balances, not Roth IRAs, which is one reason a Roth pot is useful for controlling taxable income in your seventies.