Safe Withdrawal Rate Explained

A safe withdrawal rate (SWR) is a planning assumption: the portion of your investment portfolio you withdraw each year, under a stated retirement horizon, portfolio and spending rule. It is the bridge between the portfolio you have built and the lifestyle you want to fund. The word "safe" describes how the rate was tested, not a promise about your own retirement.

The most famous benchmark is the 4% rule. In its classic form, you withdraw 4% of your portfolio in the first year of retirement, then increase that dollar amount by inflation every year after. That benchmark comes from historical tests of US stock-and-bond portfolios over 30-year periods, described in the next section.

The inverse of your withdrawal rate is your savings multiple. At 4%, you need 25× your annual expenses invested. At 3.5%, you need roughly 28.6×. At 3%, you need 33.3×. These multiples are why withdrawal-rate discussions sit at the center of every FIRE plan. You can model your own target with the Financial Independence Target Calculator.

The 4% rule traces back to two pieces of research. In 1994, financial planner William Bengen analysed every 30-year rolling retirement period in his historical dataset, testing how much a retiree could withdraw without depleting a portfolio. His historical tests used common stocks combined with intermediate-term U.S. Treasury securities. With a 50/50 mix, a 4% starting withdrawal adjusted for inflation lasted at least 30 years in every period he tested — including retirements beginning just before the 1929 crash and the stagflation of the 1970s. Bengen also argued for equity allocations between 50% and 75%, with a preference toward the upper end of that range.

Source: Bengen (1994), “Determining Withdrawal Rates Using Historical Data”.

In 1998, three professors at Trinity University — Cooley, Hubbard, and Walz — published the study that gave the rule its popular name: the Trinity Study. They tested five discrete stock-and-bond mixes (100/0, 75/25, 50/50, 25/75 and 0/100), multiple withdrawal rates, and multiple retirement lengths, reporting the historical success rate of each combination.

For a 50/50 portfolio combining the S&P 500 with long-term corporate bonds, a 4% inflation-adjusted withdrawal survived 95% of the rolling historical 30-year periods tested. Results vary with the asset mix, bond series, time period and methodology, so figures from different studies are not interchangeable.

Source: Pfau (2015), “Sustainable Retirement Spending with Low Interest Rates: Updating the Trinity Study”.

In these historical tests, "success" means the portfolio still had a positive balance at the end of the tested period. It does not mean the starting principal was preserved, and a historical success rate is not the same as the probability that a retirement beginning today will succeed. Read more in The 4% Rule Explained.

The 4% rule is a historical observation, not a contract. It tells you what would have worked in the past. It does not promise anything about the future. Several factors can push your safe rate above or below that famous benchmark.

• Retirement length. The original study focused on 30 years. A 45-year-old retiree facing 50 years needs a more conservative starting point.
• Market valuations. Starting retirement when stocks are expensive and bond yields are low raises the risk that future returns will be below historical averages.
• Geography. The most-cited safe-rate evidence is based on US market history. Pfau's 2010 study of 17 developed countries over 1900–2008 found materially different historical sustainable withdrawal rates across those markets. The study used each country's own domestic stocks, bonds and bills over 30-year retirement periods, so its results describe domestic market histories rather than one withdrawal rate for everyone outside the US. It did not directly test a globally diversified portfolio.
  Source: Pfau (2010), “An International Perspective on Safe Withdrawal Rates: The Demise of the 4 Percent Rule?”.
• Fees and taxes. The classic withdrawal-rate research does not automatically account for your individual investment costs or tax situation. Investment and advisory fees reduce the returns available to support withdrawals, while taxes can reduce the amount of a gross withdrawal that is available to spend. The effect depends on your costs, account types and jurisdiction; there is no fixed conversion from a fee percentage to a withdrawal rate.
• Spending flexibility. Real retirees adjust. The rule assumes you do not, which makes it more conservative than necessary for adaptable people.

Treat 4% as a planning anchor, not a promise. The real art is figuring out where your personal number lands given your horizon, portfolio, and ability to adapt.

One of the most underappreciated risks in retirement planning is valuation risk: the danger that you retire when markets are expensive and future returns are likely to be lower than historical averages.

Researchers have found that starting valuations — measured by metrics like the cyclically adjusted price-to-earnings (CAPE) ratio — are a strong predictor of safe withdrawal rates over the next 30 years. High CAPE ratios at retirement have historically been followed by lower safe rates. Low CAPE ratios have allowed higher starting withdrawals.

Bond yields matter too. When 10-year Treasury yields are low, the bond portion of a portfolio contributes less income and growth. That drags down the overall portfolio's ability to sustain withdrawals. A retiree starting in a high-valuation, low-yield environment faces a narrower margin of safety than the historical average suggests.

Sequence-of-returns risk is one of the most important risks facing a fixed withdrawal plan. Two retirees can earn the same average annual return and still end up with very different outcomes — purely because of the order those returns arrive, once withdrawals are being taken.

Take a deliberately simple three-year illustration. Anna and Ben each start with $1,000,000 and withdraw $40,000 at the end of each year. Both earn the same three annual returns — −15%, −15% and +20% — but in opposite order. The illustration ignores inflation adjustments to keep the arithmetic transparent.

Anna gets the losses first: her balance falls to $810,000, then $648,500, then recovers to $738,200.

Ben gets the gain first: his balance rises to $1,160,000, then falls to $946,000, then $764,100.

Both earned the same returns and the same average return, yet Ben ends roughly $25,900 ahead. With no withdrawals at all, both portfolios would finish at exactly $867,000 — the order only matters once money is being taken out. Anna sold a larger share of her portfolio at depressed prices in the first two years, and those shares were no longer invested for the recovery, leaving her on a lower path.

This is why the early years of retirement receive so much attention. A poor sequence while withdrawals are being taken generally matters more early than late, though there is no fixed point at which the risk ends — it fades gradually as the remaining horizon shortens. The risk is not the average return; it is the timing relative to withdrawals.

Defences include cash buffers, bond tents, flexible spending rules, and part-time income in the first few years. We cover the mechanics in depth in Sequence of Returns Risk Explained.

The classic 4% rule is rigid: withdraw the same inflation-adjusted dollar amount every year regardless of what markets do. Flexible withdrawal strategies instead let spending respond to portfolio performance. That flexibility changes the trade-off between income stability and withdrawal pressure; it does not remove the underlying risks.

Guyton-Klinger guardrails

Adjust withdrawals when the current withdrawal rate drifts outside preset bands: cut spending after the rate rises well above its starting level, and raise it after the rate falls well below. In Guyton and Klinger's 2006 Monte Carlo simulations, the full set of decision rules supported higher maximum initial withdrawal rates under specific assumptions. The headline results used a 40-year horizon, portfolios with at least 65% equities, and rules that allowed spending cuts and skipped inflation increases. Those are model results, not a recommended starting rate, and real income can become more variable.

Source: Guyton & Klinger (2006), “Decision Rules and Maximum Initial Withdrawal Rates”.

Dynamic SWR

Recalculate the rate each year from the remaining portfolio, remaining horizon and your return assumptions, and apply that rate to the current balance. Because the withdrawal is tied to the current portfolio value, the dollar amount automatically falls when the portfolio falls and rises when it rises. That is a mechanical property, not a guarantee that the resulting income will be adequate or sustainable for your needs. Spending can fall substantially after a bad year, so the approach requires comfort with variable income.

Cash bucket strategy

Hold a cash or short-bond reserve for near-term expenses and refill it from investments when conditions allow. A reserve can reduce how much you need to sell at depressed prices, but it does not eliminate sequence risk: a prolonged downturn can exhaust it, and cash held aside participates less in market growth. The evidence does not establish one universally correct reserve size.

Bond tent / glidepath

Shift to a more conservative allocation around retirement, then gradually increase the equity share again. Rising-equity glidepaths have been studied as a way to reduce exposure to poor returns around the start of retirement; in Pfau and Kitces' 2014 analysis they improved worst-case outcomes under the tested assumptions, not every outcome.

Source: Pfau & Kitces (2014), “Reducing Retirement Risk with a Rising Equity Glide Path”.

Estimate a baseline portfolio target with the Financial Independence Target Calculator using an adjustable withdrawal-rate assumption, then evaluate guardrail or dynamic approaches separately using the mechanics in this guide.

There is no single safe rate for everyone. The rate you test depends on your retirement length, portfolio, flexibility and other income sources. Here is what each common benchmark implies, together with the horizon evidence discussed below.

3% — a lower-rate scenario

A 3% starting rate implies about 33.3× annual expenses. A lower starting rate requires a larger portfolio for the same spending, which creates more planning margin if the assumptions hold. That trade-off does not make 3% universally safe or optimal.

4% — the classic 30-year anchor

A 4% starting rate implies 25× annual expenses. It is the benchmark associated with the Bengen and Trinity historical research discussed above, centered on 30-year retirement periods. Those historical results are evidence about the periods tested, not a recommendation for a particular retirement.

5% — a higher-rate scenario

A 5% starting rate implies 20× annual expenses. A higher starting rate reduces the portfolio required for the same first-year spending but leaves less margin if returns, inflation or the retirement horizon differ from the assumptions. Whether a rate at this level is sustainable depends on the model, horizon, portfolio and spending rules being tested.

Source: Morningstar (2025), “The State of Retirement Income: 2025”.

Confusing average and starting rates

The 4% rule refers to the starting withdrawal rate in year one. Some planners mistakenly withdraw 4% of the current balance every year. That is a fixed-percentage strategy, and it produces very different — and much more volatile — income.

Ignoring taxes and fees

The withdrawal rate is a gross portfolio figure. Taxes can reduce the amount available to spend, while investment and advisory fees reduce the returns supporting the portfolio. The gap between a gross withdrawal and after-tax spending depends on account type, investment costs and the tax rules where you live, so it cannot be reduced to one universal percentage. Model the after-tax spending you need, then work back to the gross withdrawal and portfolio assumptions required to support it.

Using 30-year data for 50-year retirements

The most widely cited Bengen and Trinity results were tested over 30-year horizons. Retiring at 40 and blindly applying a 30-year safe rate ignores the additional risk posed by two extra decades of inflation and market uncertainty.

Failing to plan for sequence risk

Many retirees assume diversification removes sequence risk. It does not. A diversified portfolio can still fall sharply early in retirement. Holding a cash or short-bond buffer and using a flexible spending rule can reduce how much you must sell at depressed prices, which can reduce exposure to a poor sequence. Neither removes it.

Setting and forgetting

Safe withdrawal planning is not a one-time calculation. Markets change, expenses change, and health changes. Revisit your withdrawal strategy at least annually and after any major life event or market shock.