Our approach to managing retirement spending draws on the probability-of-success-driven guardrails framework developed by Derek Tharp and the research team at Kitces.com. We’d recommend reading their work in full.
Here is a fact that surprises almost everyone the first time they see it: two retirees can earn exactly the same average return over thirty years, withdraw exactly the same amount each year, and end up in completely different places, with one running short and the other leaving behind more than they started with. Same portfolio, same spending, same long-run average. The only difference is the order the returns arrived in.
That’s sequence-of-returns risk, and it is the defining financial risk of the first years of retirement. It doesn’t show up in the accumulation years, when you’re adding money and can wait out any market. It appears the moment withdrawals begin, and it quietly reshapes what “average returns” even mean for a retiree. Understanding it changes how you think about the transition into retirement: not as a finish line, but as the single most return-sensitive window of your financial life.
This piece explains the mechanics (why order matters at all, when the risk is actually concentrated, and the fact that it cuts in both directions) and then walks through how we think about managing it: not with prediction, and not with product, but with a cash reserve for stability, disciplined rebalancing, and a spending plan with pre-defined guardrails that tells you in advance exactly what would change, and when.
Why order doesn’t matter until you start withdrawing
Start with the counterintuitive baseline: for a portfolio with no money going in or out, the sequence of returns is irrelevant.
Take a portfolio that experiences a 50% loss and a 100% gain, in either order. Lose half first ($1,000,000 → $500,000), then double ($500,000 → $1,000,000): you’re back where you started. Double first ($1,000,000 → $2,000,000), then lose half ($2,000,000 → $1,000,000): same place. Multiplication doesn’t care about order. Over any stretch with no cash flows, the ending balance is identical no matter how the returns are shuffled. This is why sequence risk simply isn’t a concept during most of the accumulation years: a young saver riding out a brutal decade ends up where the arithmetic says they’ll end up, and time does the rest.
Withdrawals break the symmetry. The moment you’re selling assets to fund spending, each year’s return no longer applies to the same base. A loss early in retirement, compounded by withdrawals taken at those depressed values, permanently shrinks the capital available to participate in the recovery. The shares you sold at the bottom aren’t there to ride back up. Meanwhile the retiree who got the good years first watched the portfolio grow ahead of their withdrawals, building a cushion so large that the eventual bad years landed on a much bigger base.
The result: identical average returns, radically different outcomes. Bad-years-first is the sequence that hurts, because withdrawals turn temporary declines into permanent ones. Good-years-first banks the growth before the storm arrives. Neither retiree did anything differently. They just drew different tickets in a lottery neither of them chose to enter: the year they happened to retire.
This is worth sitting with, because it reframes what retirement risk actually is. It isn’t primarily “what will markets average over the next thirty years?” Long-run averages are surprisingly forgiving. The sharper question is: what will markets do in the handful of years right after my paycheck stops? That narrow window carries wildly disproportionate weight in the outcome, and it’s the one thing nobody can know in advance.
Two retirees earn identical average returns and withdraw identical amounts. The retiree who gets the good years first builds a cushion before the bad years arrive; the retiree who gets the bad years first sells shares at depressed values, permanently shrinking the base that participates in the recovery.
Illustrative only; not actual market data or a projection. Both paths assume identical average returns and identical annual withdrawals. Only the order of returns differs.
When the risk actually bites: a decade, not a day
The instinct, once people grasp sequence risk, is to fear a crash on the eve of retirement. But a sharp one-year drop, by itself, is usually survivable. Even a severe bear market that recovers within a couple of years does limited damage, because a moderate withdrawal rate means only a small slice of the portfolio is sold at depressed prices, and in a diversified portfolio that’s rebalanced, those early withdrawals are effectively funded from the assets that held up, not the ones that fell.
The truly dangerous pattern is slower and duller: an extended stretch of poor real returns across the first years of retirement, the grinding decade where the portfolio goes nowhere while inflation-adjusted withdrawals keep coming out. No single terrifying headline, just year after year of withdrawals eating a portfolio that isn’t growing, until the base is so diminished that even a strong recovery can’t fully repair it. A crash that snaps back is a scare; a mediocre first decade is a wound. And note the inflation qualifier: a stretch of flat nominal returns with high inflation is one of the worst sequences a retiree can draw, because spending needs ratchet up right as real portfolio growth stalls. The 1970s retiree wasn’t ruined by a crash; they were ground down by inflation-eroded returns and rising withdrawal needs, year after year.
This is also why the early years dominate and the later years barely matter. A terrible market in year 25 of a 30-year retirement lands on a portfolio that has (in most sequences) already grown well ahead of the remaining spending need, with only a few years of withdrawals left to fund. The same market in years 1 through 5 hits when the remaining spending liability is at its maximum and the portfolio has had no chance to build a cushion. Sequence risk isn’t spread evenly across retirement. It’s front-loaded, concentrated in roughly the first decade, and it decays from there.
The practical takeaway: the years immediately before and after your retirement date form a risk window that deserves specific, deliberate management, different from how you invested at 45 and different from how you’ll be positioned at 80.
The part almost nobody mentions: it cuts both ways
Sequence risk has a mirror image that gets far less attention: a favorable early sequence doesn’t just avoid disaster. It compounds into surpluses that can dwarf the downside scenarios.
Think about what conservative withdrawal planning actually implies. Spending rules are calibrated to survive the worst historical sequences; that’s the entire logic of starting with a modest withdrawal rate. But the worst sequences are, by definition, rare. In most of the outcomes, the retiree who spends cautiously enough to survive the bad draw ends up with a good or ordinary draw instead, and the same compounding that punishes early losses now works in reverse, growing the portfolio far beyond what the spending plan requires. The retiree planned for the storm and got sunshine, and the caution that was essential insurance in the bad scenario becomes, in the good one, a lifetime of underspending: trips not taken, gifts not given, an estate far larger than intended by a person who lived more carefully than they ever needed to.
This is the deepest reason we believe a static retirement spending plan (pick a number, adjust for inflation, never revisit) is the wrong tool regardless of which sequence shows up. Set it low enough to survive the worst case, and you’ll almost certainly underspend your one retirement. Set it high enough to enjoy the good case, and a bad draw can sink you. No fixed number is right for both futures. The answer isn’t a better number. It’s a plan designed, from day one, to adjust in both directions.
What we don’t rely on
Before walking through what we actually do, it’s worth naming two things we deliberately don’t lean on.
We don’t rely on prediction. The market environment of your first retirement decade is unknowable in advance, and strategies that depend on forecasting it (shifting in and out of markets, or timing the entry into retirement around a market call) replace one unmanageable risk with another. Valuation measures can tell you something about whether the range of likely outcomes ahead is better or worse than average, and that’s genuinely useful context for setting a starting withdrawal rate conservatively or generously. But context is not a forecast, and we don’t build plans that require the crystal ball to be right.
We don’t rely on complexity for its own sake. There’s an entire industry of elaborate liquidation schemes: rule systems dictating which asset gets sold in which market condition. Analyzed carefully, most of the mechanical benefit these rules claim is already delivered by ordinary disciplined rebalancing. A portfolio rebalanced back to target allocations automatically sells what has run up and buys what has fallen. In a down market, that means withdrawals are effectively funded by the assets that held their value while the rebalancing process is quietly buying more of what’s depressed. The machinery that protects against selling stocks at the bottom isn’t exotic. It’s the discipline of rebalancing itself, applied consistently, especially when it feels hardest to do.
How we actually manage it
Our approach rests on three pieces that work together: a cash reserve for stability and peace of mind, a rebalanced total-return portfolio doing the mechanical work, and, most importantly, a dynamic spending plan with guardrails defined in advance.
The cash reserve: bought for peace of mind, honestly
We keep a deliberate cash reserve for clients in and near retirement, typically enough to cover a meaningful stretch of spending needs alongside other reliable income. And we’re honest about why: the largest benefit is behavioral, not mathematical.
On a spreadsheet, holding cash is a drag; over long periods it’s outgrown by nearly everything else, and research from the Kitces team testing cash-buffer strategies against actual market history has found that, purely as a returns strategy, they hurt more often than they help. The drag of the cash outweighs the benefit of avoiding down-market sales, especially since a rebalanced portfolio already avoids those sales on its own. We know that research, and we hold the reserve anyway, with eyes open, for a different reason. Retirees don’t live on spreadsheets. There is solid research showing that liquid reserves (actual accessible cash, not just wealth on a statement) measurably improve people’s sense of financial security and life satisfaction, independent of how much total wealth they have. And in a market decline, the felt difference between “our next two years of groceries are sitting in cash” and “everything we spend comes from selling into this mess” is the difference between staying with the plan and abandoning it at the worst possible moment. The most expensive mistake available to a retiree in a bear market isn’t a suboptimal withdrawal; it’s panic-selling the portfolio itself. A reserve that prevents that one mistake has paid for its drag many times over.
So we hold the reserve, we size it to the client’s sleep-at-night threshold as much as to a formula, and we don’t pretend it’s an investment strategy. It’s a stability strategy. The portfolio does the growing; the reserve does the reassuring.
The portfolio: total return, rebalanced, with the risk window respected
The investment side of sequence-risk management is less exotic than the industry makes it sound. A diversified portfolio, an allocation that respects where the client stands relative to the retirement risk window (generally more conservative through the years just before and after the retirement date than in the decades on either side), and disciplined rebalancing that mechanically sells strength and buys weakness. As covered above, that rebalancing discipline is what ensures down-market withdrawals are effectively sourced from what held up. Which account those withdrawals come from (taxable, tax-deferred, or Roth) is its own coordination problem with its own large payoff, and we’ve written separately about how we think about withdrawal sequencing across account types.
The spending plan: guardrails set in advance
The heart of our approach is the piece that addresses the both-directions problem head-on: a spending plan with pre-defined guardrails, meaning thresholds agreed on in advance that specify when spending would be trimmed and when it would be raised.
The concept (developed in the planning research world, most notably the guardrails work pioneered by Jonathan Guyton and the probability-of-success-driven refinement from Derek Tharp and Kitces.com) works like this: rather than fixing a spending number and hoping, you define a healthy zone for the plan. If markets and spending stay inside the zone, nothing changes. If a sustained decline pushes the plan’s trajectory below the lower threshold, spending is trimmed: modestly, by a pre-agreed amount, at a pre-known portfolio level. And if a favorable sequence pushes the plan above the upper threshold, spending gets a raise on the same terms. The plan breathes with the market, in both directions, according to rules written before anyone was scared or euphoric.
What we find most valuable is how the framework translates into plain dollars. Instead of abstractions, the client conversation sounds like: here’s your current spending; if the portfolio fell to roughly this level, we’d trim to this amount; if it grew to roughly that level, we’d raise to that amount. Everyone knows, in advance, where the trip-wires are and what happens at each one. When markets fall, the client isn’t wondering whether the plan is broken. They can see the portfolio is still well above the adjustment line, or they know exactly what modest trim is coming if it isn’t. The panicked phone call that starts “should we sell everything?” becomes “we’re still well above the line where we’d change anything.” That is a fundamentally different retirement experience, and it’s built not on better predictions but on better preparation.
Two properties of this approach do the heavy lifting against sequence risk. In a bad early sequence, small trims made early (a few percent, temporarily) dramatically reduce the depletion pressure of withdrawals on a diminished portfolio; flexibility is the single most powerful defense a retiree has, worth more than any product or prediction. And in a good sequence, scheduled raises solve the underspending problem, converting surplus into life actually lived rather than an accidental estate. The willingness to adjust, formalized in advance so it actually happens, is itself the risk management.
The tax overlay: a bad sequence has a silver lining
One more dimension, because a down market early in retirement isn’t only a threat. For the tax plan, it’s an opening.
The early-retirement years are often the low-tax-bracket gap years between the final paycheck and the start of Social Security and required minimum distributions, prime territory for Roth conversions. A market decline during that window makes conversions temporarily cheaper: converting shares while values are depressed moves more of the portfolio’s future recovery into the Roth, where it compounds tax-free, at a lower tax cost per share converted. The same decline that stresses the spending plan discounts the conversion opportunity. A well-built plan responds to a bad early sequence on both fronts at once, with modest spending adjustments per the guardrails and accelerated conversion work while the sale is on. Down markets in the gap years are, tax-wise, an opportunity hiding inside a problem.
Bringing it together
Sequence-of-returns risk is the reason two identical retirees with identical average returns can live completely different retirements. It exists only because withdrawals interact with the order of returns; it’s concentrated in the years surrounding the retirement date; a grinding, inflation-eroded first decade is more dangerous than a dramatic crash that recovers; and it cuts both ways, threatening ruin in the bad sequences and quiet, unnecessary underspending in the good ones. That last part shapes everything we build.
You can’t control which sequence you draw. What you can control is whether the plan was built for either draw: a cash reserve sized for genuine peace of mind, a rebalanced portfolio that mechanically avoids selling weakness, a spending plan with guardrails that specify, in advance and in dollars, what would change and when, in both directions, and a tax plan ready to treat a down market as a conversion opportunity rather than only a threat. None of it requires predicting the future. All of it requires deciding, before the future arrives, exactly how you’ll respond to it.
That’s the real answer to sequence risk: not a forecast, a framework. The retirees who navigate bad sequences well aren’t the ones who saw them coming; they’re the ones who knew, years in advance, precisely what they’d do when one showed up.
Common Questions About Sequence-of-Returns Risk
What is sequence-of-returns risk? Sequence-of-returns risk is the danger that the order in which investment returns arrive, not just their long-run average, determines a retiree’s outcome. Once withdrawals begin, losses in the early years of retirement are compounded by selling at depressed values, permanently shrinking the capital available for the recovery. Two retirees with identical average returns and identical spending can end up in completely different places purely because one experienced the bad years first.
Why doesn’t sequence risk matter before retirement? For a portfolio with no withdrawals, the math of compounding makes the order of returns irrelevant: a 50% loss followed by a 100% gain lands in exactly the same place as the reverse. Sequence risk appears only when cash flows out of the portfolio, because withdrawals taken during down years lock in losses on the shares sold. (For savers still contributing, a related timing risk exists in reverse: poor returns in the final years before retirement, when the balance is largest, can delay the retirement date itself.)
When is sequence-of-returns risk highest? It’s concentrated in the years immediately surrounding the retirement date, roughly the first decade of withdrawals. That’s when the remaining spending need is largest and the portfolio has had no chance to build a cushion. A poor market late in retirement lands on a portfolio that has usually grown well ahead of the few remaining years of withdrawals. Notably, the biggest danger isn’t a sharp crash that recovers quickly; it’s an extended stretch of poor inflation-adjusted returns across the first decade.
How do you protect against sequence-of-returns risk? The core defenses are structural, not predictive: a cash reserve covering a meaningful stretch of spending so market declines don’t force panic decisions; a diversified portfolio with disciplined rebalancing, which automatically funds withdrawals from the assets that held up while buying what’s depressed; an allocation that respects the elevated risk of the years around the retirement date; and, most powerfully, a flexible spending plan with pre-set guardrails that trims spending modestly in sustained declines and raises it in sustained good markets. Spending flexibility is the single strongest defense available.
What are retirement spending guardrails? Guardrails are pre-defined thresholds, agreed on in advance, that specify when retirement spending would be adjusted and by how much, in both directions. If a sustained market decline pushes the plan below its lower threshold, spending is trimmed by a pre-agreed amount; if a favorable market pushes it above the upper threshold, spending is raised. The approach, developed in the retirement research community, replaces a fixed withdrawal number with a plan that adapts on known terms, so a retiree always knows, in dollars, where the adjustment lines are and what happens at each one.
Is a market crash right after retiring a disaster? Usually not by itself. A sharp decline that recovers within a couple of years does limited damage at a moderate withdrawal rate, because only a small slice of the portfolio is sold at depressed prices, and with rebalancing, withdrawals are effectively funded by the assets that held their value. The genuinely dangerous pattern is a long stretch of poor inflation-adjusted returns across the first decade of retirement. A crash is a scare; a grinding decade is the real threat. A down market early in retirement can even carry a tax silver lining, making Roth conversions temporarily cheaper during the low-bracket gap years.
Fee-only fiduciary · No commissions · Always on your side of the table.