Withdrawal Rate: How Much You Can Actually Spend in Retirement
Saving enough to retire is only half the problem; spending it down without running out decades later is the harder, less discussed half. The withdrawal rate is the percentage of a portfolio spent each year in retirement, and getting it wrong in either direction either wastes years of unnecessary extra work or risks a genuinely painful shortfall late in life.
The core principle
The withdrawal rate formula is simple on its face: withdrawal rate = annual withdrawal / starting portfolio value. A retiree with a $1,000,000 portfolio who withdraws $40,000 in the first year of retirement is using a 4% withdrawal rate. Most standard retirement models then hold the dollar amount roughly constant in real terms, adjusting it for inflation each subsequent year, rather than recalculating the percentage against a portfolio balance that is constantly rising and falling with the market.
The widely cited 4% guideline traces back to historical research testing rolling 30-year retirement periods using US market returns and a portfolio mix of roughly 50% to 75% stocks. Across nearly all of those historical 30-year windows, an initial 4% withdrawal, adjusted upward for inflation every year after, did not exhaust the portfolio before the 30 years were up. That is a genuinely useful anchor for retirement planning, but it is an empirical result from a specific historical sample, not a law of physics: it assumes a 30-year time horizon, a specific stock-to-bond mix, and that future US markets behave broadly like the historical sample the research was built on.
A subtlety the simple percentage hides entirely is sequence-of-returns risk: the order in which gains and losses arrive matters enormously to a portfolio that is simultaneously being withdrawn from, even when the average return over the full period is identical. A portfolio earning the exact same average return as another can end up with a dramatically different balance decades later purely because the bad years happened to fall early rather than late in retirement.
It is worth noting what the withdrawal rate research does and does not claim. It does not claim that 4% is the mathematically maximum safe rate, nor that every historical 30-year window produced an identical ending balance; some historical starting points left a retiree with several times their original balance after 30 years, while others left the portfolio nearly, but not quite, exhausted. The 4% figure specifically describes the rate that survived the worst historical 30-year stretch in the sample studied, which is precisely why it functions as a conservative planning anchor rather than an average expected outcome.
How the math works
Example 1: the basic calculation and its inflation adjustment. A retiree begins with $1,200,000 and uses a 4% starting withdrawal rate: $1,200,000 x 0.04 = $48,000 in year one. If inflation runs 3% that year, the year-two withdrawal is adjusted upward to $48,000 x 1.03 = $49,440, regardless of what the portfolio's market value actually did that year. This is the defining feature of the standard model: the dollar amount, not the percentage, is what gets held roughly constant in real terms year to year, which is exactly why a sustained bad market stretch early on can push the effective withdrawal percentage against the portfolio's shrinking balance well above the original 4%.
Example 2: sequence-of-returns risk with identical average returns. Two retirees each start with $1,000,000 and withdraw $40,000 at the start of each year, before that year's market return is applied, over a simplified three-year stretch. Retiree A experiences returns of −20%, +8%, +8% in that order; Retiree B experiences the same three returns in reverse order: +8%, +8%, −20%. Both sets of returns average exactly (−20% + 8% + 8%) / 3 = −1.33%. Retiree A: withdraw $40,000 leaving $960,000, apply −20% to get $768,000; withdraw $40,000 leaving $728,000, apply +8% to get $786,240; withdraw $40,000 leaving $746,240, apply +8% to end at $805,939. Retiree B: withdraw $40,000 leaving $960,000, apply +8% to get $1,036,800; withdraw $40,000 leaving $996,800, apply +8% to get $1,076,544; withdraw $40,000 leaving $1,036,544, apply −20% to end at $829,235. Despite identical average returns and identical withdrawals, Retiree A ends roughly $23,300 poorer than Retiree B, purely because the bad year landed first. Over a real 30-year retirement, this same mechanism can compound into a difference of hundreds of thousands of dollars, or the difference between a portfolio lasting the full retirement and running out early.
How it shows up in real portfolios
For a traditional retiree following a fixed percentage rule mechanically, the practical risk shows up precisely during the first five to ten years of retirement: a market downturn early on, combined with continued fixed-dollar withdrawals, can permanently impair the portfolio's ability to recover, even if the market fully rebounds later, because the withdrawals during the downturn were made against a smaller base and locked in a permanently lower share count. This is the specific mechanism behind sequence-of-returns risk, and it is why many retirement planners recommend building in flexibility, sometimes called guardrails, that trims spending somewhat during a confirmed bad stretch rather than mechanically withdrawing the same inflation-adjusted dollar amount regardless of market conditions.
A high-earning-professional scenario shows the horizon problem clearly: a physician or executive pursuing early retirement in their late 40s or early 50s is not planning around the 30-year window the original 4% research was built on, but a horizon that could stretch 40 years or more. Because a longer horizon gives sequence-of-returns risk more years to compound and more total time for the portfolio to be tested, early retirees commonly adopt a more conservative starting rate, often in the 3% to 3.5% range, or build in explicit spending flexibility, to compensate for the extra decade or two the original research never tested.
Retirees starting withdrawals at a moment of unusually high market valuations face a related, though distinct, concern: a portfolio that begins its withdrawal phase at elevated valuations has historically shown a wider range of subsequent outcomes than the average across all historical starting points, which is one of the more evidence-based arguments for treating a single fixed percentage as a planning anchor rather than a guarantee calibrated to every possible starting environment.
Actionable breakdown
- Before settling on a withdrawal rate, consider:
- Your actual expected retirement horizon, not a default 30 years.
- Current market valuations relative to long-run historical norms.
- Your stock-to-bond mix relative to the mix the original research used.
- Whether you have flexibility to reduce spending in a bad early stretch.
- Watch for these red flags:
- Treating 4% as a guarantee regardless of your specific starting conditions.
- Planning a 40-plus year retirement around 30-year historical research.
- A rigid budget with zero flexibility to cut spending in a downturn.
- Ignoring how much a bad first five years can permanently impair the portfolio.
- Treat 4% as a starting point, not a guaranteed safe number.
- Lower the starting rate for retirements longer than 30 years.
- Revisit spending yearly using guardrails rather than a fixed formula.
- Keep genuine flexibility to reduce spending during a confirmed bad stretch.
Social Security, pensions, and other guaranteed income sources change this calculation meaningfully for many retirees, since a withdrawal rate applied only to the portfolio, ignoring a separate, inflation-protected income floor already covering a portion of expenses, tells an incomplete story. A retiree whose guaranteed income already covers half of annual spending can reasonably tolerate more variability in the portfolio-funded half than the standard research, built around a scenario with no such floor, would suggest on its own.
Common pitfalls
- Treating the 4% guideline as guaranteed safe regardless of market conditions at retirement, ignoring that starting valuations meaningfully affect historical outcomes.
- Withdrawing a fixed inflation-adjusted dollar amount with no floor or flexibility, which can deepen losses sharply during a bad early market stretch.
- Planning for a fixed 30-year retirement when an early retiree may genuinely need the portfolio to last 40 years or more.
- Ignoring sequence-of-returns risk entirely, assuming that hitting a target average return is sufficient regardless of the order those returns arrive in.
Related concepts
For the mandatory counterpart to voluntary withdrawals in tax-deferred accounts, see required minimum distribution. For the broader early-retirement framework built around this rate, see FIRE. For the full set of strategies used to manage this risk in practice, see the guide on withdrawal strategies and the guide on FIRE.
The bottom line
A withdrawal rate is a planning anchor built from a specific historical sample, and a real retirement plan needs both a conservative starting point and genuine flexibility built in around it.