The Safe Withdrawal Rate Number Is Solving The Wrong Problem

Executive Summary

Every year, the retirement planning industry waits for an updated safe withdrawal rate. Morningstar puts this year’s number at 3.9 percent, up from last year, and the debate that follows treats the decimal point like it’s the whole plan. Bump the number up, a client can spend more. Bump it down, they spend less. Either way, the number is treated as the thing that decides whether a retirement lasts thirty years.

It doesn’t decide that. What decides it, more than the withdrawal rate itself, is what the market does in the first five to ten years after a client stops working. Get that part wrong and no starting percentage saves the plan. Get it right and the plan has room to be generous. The industry keeps refining the number and mostly ignoring the years that actually determine the outcome.

Carol retired at sixty-two with $1.1 million and a plan built around a 4 percent withdrawal rate. The number felt safe, and on paper it was, backed by decades of research showing a portfolio built that way survives most thirty-year stretches. Fourteen months into retirement, the market dropped 18 percent, and her advisor got the call every advisor eventually gets. Should I stop taking my draw. Should I have never started.

The honest answer wasn’t about the 4 percent. It was about the fact that nobody had walked through what happens if the drop shows up in year one instead of year fifteen.

The Withdrawal Rate Number Changes Every Year, And That Misses The Point

Morningstar’s 2026 State of Retirement Income report puts the baseline safe withdrawal rate at 3.9 percent for a retiree using fixed, inflation-adjusted spending over a thirty-year horizon, up from the year before, an increase driven by improved capital markets assumptions rather than anything a retiree did differently. For a retiree willing to flex spending with the market, using a guardrails approach, the same research allows as much as 5.7 percent, according to a summary of the report. William Bengen, whose original research produced the 4 percent figure everyone still argues about, has since revised his own estimate upward, to 4.7 percent, in his book A Richer Retirement.

None of these numbers are wrong. They’re just answering a question that assumes the retiree experiences something close to an average sequence of market returns. Almost nobody does. Two retirees can earn the exact same average return over thirty years and end up in completely different places, one running out of money at eighty-four, the other leaving an estate, purely because of the order the returns arrived in. The safe withdrawal rate literature has known this for decades. The annual number that gets published and argued over is a headline that mostly skips it.

The industry is already moving this direction in practice. Advisors are increasingly treating the 4 percent rule as a starting reference rather than a plan, replacing a single upfront percentage with an approach that gets monitored and adjusted as markets and a client’s actual life move, which is closer to what the research has supported all along.

What Actually Decides Whether A Retirement Plan Survives

Wade Pfau’s research on what he calls the lifetime sequence of returns quantifies this directly. Using Monte Carlo simulation and regression analysis across thousands of overlapping thirty-year retirement periods, Pfau found that the compounded return in the first ten years of retirement explains 77 percent of the variation in the final retirement outcome, while the return over the remaining twenty years explains a small fraction of that. The years right around retirement carry almost the entire weight.

That’s not a modest tilt. Three-quarters of the outcome comes down to a decade most retirees don’t think to plan around, because most retirement conversations are still organized around a single number instead of a timeline. A client who retires into a bull market and a client who retires into a bear market can start with the same balance, the same withdrawal rate, and the same investment policy, and finish decades apart, for reasons that had nothing to do with either client’s decisions.

Required Minimum Distributions Complicate A Plan That Was Never Just About A Percentage

Sequence risk isn’t the only thing layered on top of the withdrawal rate conversation. Required minimum distributions start at age 73 under current law, and they don’t ask whether the market cooperated that year. A retiree forced to distribute a set percentage of a tax-deferred account during a down year is selling into the same decline sequence risk already punishes, and missing the distribution entirely carries a 25 percent excise tax on the shortfall, reduced to 10 percent if corrected within two years. A withdrawal strategy that only accounts for spending, and not for the RMD schedule arriving on its own timetable regardless of market conditions, is missing a second, mandatory version of the same risk.

Delaying Social Security interacts with this too, in the opposite direction. Every year a client waits past full retirement age adds 8 percent to the eventual benefit, up to age 70, which means a client who delays from 67 to 70 collects 24 percent more for the rest of their life. That increase is fixed by statute rather than by markets, and it’s one of the few levers in a retirement plan that isn’t exposed to sequence risk at all. Advisors who spend more time optimizing an asset allocation than they spend on the claiming decision are optimizing the smaller lever.

A Three-Signal Way To Check A New Retiree’s Actual Exposure

None of this requires abandoning the withdrawal rate conversation, it requires adding three questions to it before a client’s first year of distributions begins.

Signal one: how much of the portfolio is in cash or short-term bonds against the next one to two years of spending. A client with enough set aside to cover a couple of years of withdrawals never has to sell equities into a downturn, which is the mechanism that actually turns sequence risk into permanent damage.

Signal two: is the spending plan fixed or flexible. Vanguard’s research on dynamic spending shows a retiree using a flexible approach, one that lets annual spending move within a ceiling and a floor around the prior year’s amount rather than rising automatically with inflation, starting around 4 percent and adjusting the following year’s number up to a 5 percent ceiling or down to a 2.5 percent floor depending on how the portfolio performed. That flexibility is what lets a plan absorb a bad first decade instead of locking in a fixed draw against a shrinking balance.

Signal three: what happens to the plan if the first five years look like the worst five years on record. Most retirement conversations are still run on an average-return assumption instead of that one.

A retiree who passes all three checks can often handle a higher withdrawal rate than the annual headline number suggests. A retiree who fails all three should probably be planning around something more conservative than 3.9 percent, regardless of what this year’s report says.

Running This Conversation In The Actual Client Meeting

The three signals work better as a conversation than a worksheet, mostly because the client rarely brings up sequence risk on their own, they bring up the number they read about in an article. The opening question that gets somewhere is not “what withdrawal rate are you comfortable with,” it’s “if the market drops 20 percent in your first year retired, what do you want to happen to your spending.” Most clients haven’t thought about it in those terms, and the answer tells you immediately whether the plan in front of them has a floor or just an assumption.

From there, the cash buffer conversation is concrete rather than theoretical. A client with eighteen months of spending in short-term bonds can be shown, in the same meeting, exactly which account funds year one and year two if equities are down double digits, and that specificity does more to calm a retiree than a probability-of-success number ever does. The RMD schedule and the Social Security claiming decision belong in that same meeting, because a client who is delaying benefits to age 70 is, in effect, self-funding three extra years of spending from the portfolio, which changes how much cash buffer that period actually needs.

None of this replaces the withdrawal rate conversation. It just makes sure the conversation happens before the first bad year, not during it, which is the only point in the process where any of it is still useful.

The Number Still Matters, It’s Just Not The Whole Conversation

Carol’s plan survived, not because 4 percent turned out to be the right number, but because there was enough set aside in short-term bonds that nothing had to be sold during the 18 percent drop, and because her spending plan had a floor built in rather than a fixed draw that assumed the market owed her an average year. The withdrawal rate was the same number it had been on day one. What changed the outcome was the planning that happened before the drop ever showed up.

Our goal is to make sure that planning happens before the call, not during it. Carol’s advisor got lucky that the answer was already built into the plan. Most advisors only find out whether theirs is built in when a client is already on the phone.

Withdrawal rate is one piece of retirement income planning. We write about the rest of it, claiming strategy, RMDs, income volatility, on our retirement planning page.