The Drawdown Stress Test — Will Your Portfolio Survive a 2008?

The 4% Rule was designed for a 30-year retirement starting at 65. Apply it to a 45-year-old retiree and it understates the failure rate by a factor of three.

William Bengen published the original 4% Rule in 1994 using historical U.S. data from 1926 through 1976 and a 30-year retirement horizon. That sample window and that horizon are the entire reason the number is 4%. Change either input — say, by starting retirement at 45 instead of 65, or by extending the window through the lost decade of 2000–2009 — and the safe rate falls. The math hasn't been hidden; it's been quietly absorbed into a slogan that survives in conversations decades after the assumptions behind it stopped applying to the people repeating it.

This page is the simplest honest version of the calculation. Pick a portfolio, pick a withdrawal, pick a historical starting year, and the sandbox to the right shows you year-by-year what would have happened. Move the starting-year lever to 2000 and the 4% withdrawal that Bengen's sample comfortably supported looks much more fragile. Move it to 1980 and the same withdrawal feels almost wasteful. Same average return, opposite outcomes. The story below explains why.

What is the 4% Rule, really?

Stated precisely: withdraw 4% of your starting portfolio balance in year one, then increase that dollar amount by inflation each subsequent year, regardless of what the portfolio is doing. A $1,000,000 portfolio in year one supports a $40,000 withdrawal; if inflation runs 3%, year two's withdrawal is $41,200, and so on. The protocol holds the lifestyle constant, not the withdrawal percentage. By year fifteen, after a string of strong markets, you might be drawing 2% of your balance. After a bad decade, you might be drawing 8%. Either way, the dollar amount is locked.

This matters because two strategies that sound similar are not. "I'll withdraw 4% of my balance each year" is a different strategy — and almost always a safer one — because it lets withdrawals shrink with the portfolio. Bengen's rule does not. It is, by design, rigid: he wanted a protocol a retiree could follow without recalculating each year, and the trade-off for that simplicity is concentrated tail risk when markets misbehave early.

Two more modern alternatives address that trade-off explicitly. Guyton-Klinger introduces guardrails: if the current withdrawal rate climbs more than 20% above the initial rate after a drawdown, you cut spending by 10% temporarily. Variable Percentage Withdrawal (VPW) recalculates the safe withdrawal each year as a function of remaining horizon and current balance. Either typically buys 50–100 basis points of safe withdrawal in exchange for tolerating income volatility. The full simulator at /simulators/financial-projections supports all three modes; the sandbox here intentionally locks you into Bengen's protocol so the sequence-risk lesson is uncluttered.

balance(t) = balance(t-1) · (1 + r_t) − withdrawal_t
withdrawal_t = withdrawal_0 · (1 + inflation)^(t-1)

Sequence of return risk: the silent killer

Two retirees retire with identical $1,000,000 portfolios and an identical $40,000 inflation-adjusted withdrawal. Both experience the same set of annual returns over thirty years. The only difference is order. Retiree A gets the bad years first. Retiree B gets them last. Even though the geometric mean return is identical, Retiree A ends with an exhausted portfolio and Retiree B ends with millions to spare. That asymmetry — the same average producing wildly different outcomes depending on order — is sequence of return risk.

It exists because withdrawals lock in losses. A 30% market drop in year one is followed by a withdrawal that takes another $40,000 out of an already-reduced base. The remaining balance now has to compound back from a smaller number; even if subsequent returns are great, the portfolio never fully catches up. By contrast, a 30% drop in year twenty-five matters far less: the portfolio is either close to the finish line or has compounded so far ahead that the haircut doesn't bite. Order doesn't matter for an investor who isn't withdrawing. For a retiree, it's almost everything.

Try it in the sandbox. Set the starting year to 2000 — the dot-com crash followed by the 2008 crash, two recessions in the first decade. Then set it to 1980 — the start of a generational bull market that ran most of the way to 2000 before any meaningful correction. Same portfolio. Same withdrawal. The 2000-start retiree depletes; the 1980-start retiree dies rich. The arithmetic mean of historical returns from 1980–2009 is, in fact, very close to the mean from 2000–2029. Only the order changed.

The 30-vs-50-year asymmetry

Bengen's sample ended in retirements that began as late as 1976. A retiree starting in 1976 with a 30-year horizon was finished by 2005 — before the 2008 crash mattered to them. A 45-year-old retiring today with a 50-year horizon will encounter not one but two or three full market cycles before drawing their last withdrawal. Two implications follow.

First, the failure rate doesn't grow linearly with horizon. Doubling the retirement length doesn't double the failure probability; for the 4% Rule against U.S. historical data, the 30-year failure rate is roughly 5%, the 40-year failure rate is roughly 15%, and the 50-year failure rate climbs to 18–25% depending on the specific dataset and rebalancing rules. The shape is roughly cubic in the early decades and then plateaus. Bengen's "safe" turns into a one-in-five coin flip at the 50-year horizon. That is the "factor of three" understatement.

Second, withdrawal rate compresses dramatically. The rate that produces the same failure probability at 50 years as 4% does at 30 is closer to 3.0–3.3%. Translation: an early retiree needs roughly 25–30% more capital to support the same lifestyle the traditional retiree supports at 4%. A $1,000,000 portfolio that funds $40,000/year for a 65-year-old funds $30,000–33,000/year for a 45-year-old at equivalent risk. That delta is the part of the FIRE conversation that often goes unsaid until someone runs the numbers.

Counter-cases: when 4% (or more) is fine

The argument above is one-sided on purpose; the sandbox lets you build a one-sided case in the other direction just as easily. There are real conditions under which the 4% Rule is conservative, even for early retirees. Three of them.

You have flexible spending. Bengen's protocol assumes you spend the inflated dollar amount every year no matter what. Real retirees don't. In a recession year, most households cut discretionary travel, defer the kitchen remodel, eat in. Even a 10% reduction in spending during the worst three years of a sequence is often enough to rescue a portfolio that would otherwise have failed. Variable strategies institutionalize this; informally, most retirees already do it. The model fails worse than reality.

You have meaningful future income. Social Security at 67 or 70, a pension, rental income, royalties, or a partner who keeps working part-time all reduce the portfolio withdrawal needed in those years. A retiree whose Social Security covers 60% of expenses starting at 67 is, mathematically, only stress-testing the bridge to 67. The 50-year horizon shrinks to a 22-year horizon, which is well inside Bengen's safe zone.

You'll work again if you have to. The hardest case to model is human optionality. A 45-year-old retiree who would, in a genuine crisis, return to consulting work for two or three years isn't running a fixed-withdrawal problem at all. They're running a stochastic cash-flow problem with an embedded option. The option's value is enormous and uncountable; ignoring it can make a plan look fragile when the human behind it is robust.

What to do today

Four concrete actions, ranked roughly by impact per hour of effort.

1. Re-test with your real spending. The default $40,000 withdrawal is illustrative. Plug your actual projected retirement spending into the sandbox, pick the three worst historical starts (1929, 1966, 2000), and see where you land. If two of three survive, your plan is roughly Bengen-safe; if only one survives, your headline withdrawal rate is too high.

2. Hold a cash buffer. One-to-three years of expenses in a high-yield savings account or short-duration bonds, drawn down preferentially during equity drawdowns, removes the worst sequence-risk dynamic: forced selling at the bottom. The opportunity cost is modest (a few hundred basis points on the cash portion). The downside protection is substantial.

3. Build a Roth conversion ladder. Early retirees who want access to pre-tax money before 59½ without penalty should understand the five-year seasoning rule on Roth conversions. The Roth Conversion Playbook covers it in detail. The relevant point here: a conversion ladder gives you a second source of withdrawal flexibility, which itself cushions sequence risk.

4. Run Monte Carlo against your actual plan. The sandbox here runs one history at a time. The Pro tier of the simulator runs 10,000 randomized histories and reports the probability your plan survives — a more honest answer than any single historical starting year can provide.

Where to go from here

If your spending is below Bengen's 4% in real terms and your horizon is under 35 years, the historical record is on your side and there isn't much more to do. If either condition fails, the right next step depends on which one. For long horizons, the full simulator chains deterministic and Monte Carlo modes against your real accounts. For unusually high projected spending, the Zero-Percent Harvest and Roth Conversion Playbook blueprints walk through the two largest tax-arbitrage levers an early retiree has. And if you'd rather have a model talk to you about all of this with your actual numbers, the AI Advisor reads your full profile and proposes a year-by-year drawdown plan.