Sequence of Returns Risk: What a Bad First Five Years Does to a $3 Million Retirement, and the Four Defenses That Work

Written by Hazel Secco, CFP®, CDFA®

Estimated reading time: 17 minutes

“Retiring Soon and Worried About Sequence of Returns Risk.” “Retiring into a recession.” “When does sequence risk kick in?” Those are three Bogleheads thread titles from the last two years, and they come from people who have done everything right: saved, diversified, reached the number. What they are asking is whether the number survives if the market falls the year after they stop working.

Sequence of returns risk is the danger that poor returns arrive early in retirement, while you are withdrawing, so that the portfolio is smaller when the recovery comes. Two retirees can earn the same average return over thirty years and end up millions of dollars apart, purely because of the order. For a woman with $3 million and a retirement that may run three decades, it is the single largest risk in the first ten years, and it is also the most manageable, because every defense against it is a decision she controls.

This article replays the real returns from 2000 through 2025 on a $3 million portfolio, shows what a 4 percent and a 5 percent withdrawal did in that sequence, and then tests four defenses against the same history: a guaranteed spending floor, spending guardrails, a rising equity glide path, and a cash reserve. One of the four did not help, and I will show you why.

What is sequence of returns risk?

Sequence of returns risk is the effect of the order of investment returns on a portfolio that is being drawn down. During accumulation, order does not matter: a 20 percent loss followed by a 25 percent gain leaves you where you started, whichever comes first. Once withdrawals begin, a loss in year one is compounded by the money you took out at the bottom, and the portfolio that recovers is smaller than the one that fell. The average return can be identical; the ending balance is not.

The cleanest way to see it is to run the same returns forward and backward. The table below uses the actual S&P 500 and Bloomberg U.S. Aggregate Bond returns from 2000 through 2025, with 7 percent stocks and 4 percent bonds assumed for the last four years, on a 60/40 portfolio rebalanced annually. In one case the retiree starts in 2000 and absorbs three losing years immediately; in the other she gets the same 26 years in reverse, starting with 2025 and ending with 2000. Both earn 6.76 percent a year on the 60/40 mix over thirty years.

What does a bad first five years do to a $3 million retirement?

A woman who retired on January 1, 2000 with $3 million, withdrawing $120,000 a year and raising it 2.5 percent for inflation, watched the portfolio fall to $2.28 million after three years and touch $2.06 million in 2008. She never ran out; the 2009 to 2021 bull market rescued her, and she ended year 30 with about $2.77 million. The same returns in reverse order, with the good years first, would have ended at $8.96 million. Same average, same withdrawals, a $6.2 million difference from sequence alone.

$3 million, 60/40, $120,000 initial withdrawal (4%) rising 2.5% a year	After year 5	After year 10	After year 20	After year 30
Actual 2000 to 2025 order (three losses first)	$2.62 million	$2.26 million	$2.81 million	$2.77 million
Same returns in reverse (2025 first)	$3.78 million	$5.41 million	$8.02 million	$8.96 million
Constant 7% stocks / 4% bonds every year	$3.23 million	$3.44 million	$3.69 million	$3.47 million

The 4 percent withdrawal survived the worst starting decade in modern history, which is the point of the 4 percent rule. Raise the withdrawal to 5 percent, $150,000 a year, and the 2000 start runs out of money in year 25 while the reversed order still ends with $5.9 million. At 6 percent, $180,000, the money is gone in year 18. The difference between “comfortable” and “ran out at 85” in a bad sequence is not the market; it is the size of the withdrawal in the first ten years.

These figures ignore taxes and fees, and they assume she kept withdrawing on schedule through 2002 and 2008 without changing anything, which is exactly what the four defenses below are designed to fix. I use $3 million because that is where many of my clients start, and because every example in the big-brand articles uses $1 million and $40,000, which hides how the arithmetic feels when the loss is $700,000 in three years. Whether the 4 percent rule works for women with longer horizons is its own question; this article is about what happens in the first decade regardless of the rate.

How long does the danger window last?

The danger window is roughly the first ten years of retirement, with the five years before retirement and the five after being the most sensitive. William Bengen, who created the 4 percent rule, has said that when plans break, they break in the first ten to twelve years. Michael Kitces calls the final decade before retirement and the first half of retirement the “retirement red zone.” After that, a portfolio that has survived is usually large enough relative to the remaining withdrawals that later bear markets cannot sink it.

For women the window does not shrink; the horizon after it grows. On the Social Security Administration’s current period life table, a 65-year-old woman has 20.7 years of remaining life expectancy, about 35 percent of 65-year-old women reach 90, and 14 percent reach 95. A 60-year-old woman retiring in 2026 should plan for thirty-five years, which is why the 2000 replay above, ending at $2.77 million in year 30, is not the reassuring number it looks like. She needs it to last another five to ten years past that.

Defense one: a spending floor that does not depend on the market

The first defense is to shrink the amount the portfolio has to produce in the early years, so that a bad sequence hits a smaller withdrawal. The two tools are delaying Social Security and keeping some earned income. A woman with a $3,600 primary insurance amount who claims at 70 instead of 62 receives $53,568 a year instead of $30,240, inflation-adjusted for life, and that $23,000 difference is a permanent reduction in what the portfolio must cover after 70, which is when the portfolio is most likely to have been damaged by an early bear market.

Part-time or consulting income in the first three to five years does the same job on the front end. Every $40,000 of earned income in a down year is $40,000 the portfolio does not have to sell at the bottom. In the 2000 replay, a retiree who covered a third of her spending with consulting income through 2002 would have finished the decade about $160,000 ahead of the one who did not, and nearly $900,000 ahead by year thirty. I do not suggest this because anyone needs to work in retirement. I suggest it because the first five years are the only years in which a modest income changes the whole trajectory.

Defense two: guardrails instead of a fixed 4 percent

The second defense is a withdrawal rule that responds to the portfolio instead of ignoring it. Guyton and Klinger’s guardrails, published in the Journal of Financial Planning in 2006, work like this: start at 4 to 5 percent, raise the withdrawal for inflation each year, but if the current withdrawal rate rises more than 20 percent above where it started, cut the dollar amount 10 percent and skip that year’s inflation increase; if it falls more than 20 percent below, raise it 10 percent. For a 4 percent start, the guardrails sit at 4.8 and 3.2 percent.

Applied to the 2000 replay, the guardrails triggered cuts in 2003, 2009, and 2010, and raises in 2021 and 2022. The retiree ended year 30 with $6.38 million instead of $2.77 million. The price was real: she withdrew $4.1 million over thirty years instead of $5.3 million, and her year-30 income was $201,000 instead of $246,000. Morningstar’s 2025 study reaches the same conclusion from the other direction: a fixed inflation-adjusted withdrawal supports a 3.9 percent starting rate at 90 percent confidence, while a guardrails approach on a 40/60 portfolio supports 5.2 percent, because the retiree accepts that spending will move.

2000 replay, $3 million, four defenses tested one at a time	After year 10	After year 30	Total withdrawn over 30 years
No defense: fixed 4% rising with inflation, 60/40	$2.26 million	$2.77 million	$5.27 million
Guardrails: cut 10% above 4.8%, raise 10% below 3.2%	$2.42 million	$6.38 million	$4.14 million
Rising equity glide path: 40% stocks in year 1 to 70% by year 10	$2.57 million	$6.07 million	$5.27 million
Cash reserve: two years of spending held at 4%, used after down years	$2.32 million	$2.49 million	$5.27 million

Kitces has pointed out that withdrawal-rate guardrails can demand cuts of 28 to 54 percent in the worst historical sequences, which is more than most households can absorb. In practice I set the guardrails on discretionary spending, the travel and gifts and the second home, not the mortgage and the health insurance. A 10 percent cut to a $150,000 budget that has $50,000 of discretionary spending in it is a 30 percent cut to the discretionary line, and that is the conversation to have before the bear market, not during it.

Defense three: a rising equity glide path

The third defense is to hold fewer stocks at the start of retirement and more later, the opposite of the conventional advice. Wade Pfau and Michael Kitces showed in the Journal of Financial Planning in 2014 that a portfolio starting at 30 percent equities and rising to 60 percent outperformed one held at 60 percent throughout, because the low allocation protects the portfolio during the years it is most vulnerable, and the rising allocation captures growth once the danger window has passed. Their optimal paths start at 20 to 40 percent stocks and finish at 50 to 70 percent.

In the 2000 replay, a glide path from 40 percent stocks in year one to 70 percent by year ten ended at $6.07 million with the same $5.27 million of withdrawals as the base case, a $3.3 million improvement from allocation alone. Two honest caveats. The glide path would have underperformed had the good years come first, so it is insurance, not a return enhancer. And the mechanism works because you are buying stocks through the decline: by 2003 the retiree is at 47 percent equities and adding, which requires a plan written in advance and the nerve to follow it.

The practical version for a woman retiring in 2026 is what Kitces calls a bond tent: build the bond and cash allocation up in the five years before retirement, let it peak on the retirement date, and spend it down over the first five to ten years while equities are left to compound. It is the same idea as the glide path with a start date you can put on a calendar.

Defense four: the cash reserve, and what it is actually for

The fourth defense is the one everyone asks about first, and it is the one that did not add money in the test. A two-year cash reserve, $240,000 carved out of the $3 million, earning 4 percent, spent only in the year after a losing stock year and refilled in good years, ended the 2000 replay at $2.49 million, about $280,000 behind the retiree who simply held 60/40 and rebalanced. The reason is that a 60/40 portfolio rebalanced annually already sells bonds to fund withdrawals after a stock decline; the cash bucket duplicates that and adds a drag in the other 20 years. Kitces reached the same result comparing bucket strategies with plain rebalancing.

So why do I still hold one for clients? Because the spreadsheet assumes the retiree kept withdrawing on schedule through 2002 and 2008 without selling stocks in a panic, and human beings do not behave like spreadsheets. A visible cash account with eighteen months of spending in it is what lets a woman watch her portfolio fall $700,000 and not call to sell. Its job is behavioral, and it should be sized for that job: one to two years of spending, not five. Beyond that, every extra year in cash is a year of returns you are giving up to feel calm, and the glide path above does the same work at a lower cost.

Where the reserve lives matters in New Jersey and New York. Cash held inside an IRA is withdrawn as taxable income; cash in a brokerage account is not. In a down year you want the flexibility to draw from the taxable account and leave the IRA for conversion, which is a withdrawal sequencing decision made before the market falls, not after.

The silver lining: a Roth conversion in the down year

A bear market in the first years of retirement is the best Roth conversion opportunity most people will ever get. A conversion is taxed on the fair market value of what you move on the day you move it, so converting $200,000 of an IRA that was $280,000 a year earlier moves the same shares to the Roth at 29 percent less tax, and every dollar of the recovery happens tax-free. Since 2018 conversions cannot be undone, so the amount has to be chosen deliberately, but the direction is clear: the year the portfolio is down is the year to convert more, not less. This is the Roth conversion window with the market on your side.

Three constraints at your level. The 24 percent bracket ends at $201,775 of taxable income for a single filer in 2026, and conversions above it are taxed at 32 and 35 percent. Medicare IRMAA looks back two years, so a large conversion at 63 raises the Part B premium at 65. And New Jersey taxes the conversion at up to 6.37 percent with no capital-loss offset, since the state does not allow losses in one category to reduce income in another; New York taxes it above the $20,000 pension exclusion. I lay the state rules out in retiring in a high-tax state. None of these is a reason not to convert in a down year. All of them are reasons to size the conversion to the bracket and the IRMAA line rather than to the fear.

What to Do This Week

Calculate your first-year withdrawal rate. Divide the amount you will take from the portfolio next year, after Social Security, pension, and any earned income, by the portfolio value. Below 4 percent, a bad sequence is survivable on the history we have. Above 5 percent, it is not, and something in the plan has to change before the market does it for you.

Write down the guardrails and what gets cut. Pick the withdrawal rate that triggers a 10 percent cut, usually 20 percent above your starting rate, and list the discretionary items that absorb it. A plan that says “we will spend less if the market falls” is not a plan; a plan that says “we skip the second trip and the car if the rate passes 4.8 percent” is.

Set the retirement-date allocation on purpose. If you are within five years of retiring, the equity percentage on your retirement date should be the lowest of your life, and it should be scheduled to rise. Decide the starting number, the ending number, and the years in between now.

Fund the reserve from the right account. Eighteen months of spending, in a taxable account or high-yield cash, not inside the IRA. Note the date it was last refilled and the rule for refilling it.

Pre-decide the down-year conversion. Write the number you would convert if the portfolio falls 20 percent, sized to the top of the 24 percent bracket after your other income. When the year comes, you execute, you do not deliberate.

Frequently Asked Questions

What is sequence of returns risk in simple terms?

It is the risk that bad investment years come early in retirement, while you are withdrawing money, so the portfolio is smaller when the good years arrive. Two retirees with the same average return over thirty years can end up millions apart depending on which years came first. It only matters once withdrawals begin; while you are saving, the order of returns makes no difference.

How many years of cash should a retiree hold?

One to two years of spending, held outside the IRA, is enough for the job a cash reserve actually does, which is keeping you from selling stocks in a decline. In a historical test on a $3 million portfolio, a two-year reserve earning 4 percent ended thirty years about $280,000 behind plain rebalancing, so larger reserves cost real money. Five years of cash is a return sacrifice, not a safety measure.

Does the 4 percent rule still work in 2026?

Morningstar’s 2025 study puts the safe starting rate for a fixed, inflation-adjusted withdrawal at 3.9 percent over thirty years with 90 percent confidence, and 5.2 percent with guardrails on a 40/60 portfolio. William Bengen’s 2025 update raised his own historical worst-case rate to 4.7 percent, but that assumes a broadly diversified portfolio including small-cap and international stocks. For a woman with a 35-year horizon, a flexible rule beats a fixed rate.

Should I hold more or fewer stocks when I retire?

Fewer at the start, more later. Research by Pfau and Kitces found that starting retirement at 20 to 40 percent equities and rising to 50 to 70 percent produced better outcomes than a constant allocation, because the low allocation protects the portfolio in the years it is most vulnerable to a bad sequence. In a replay of the 2000 to 2025 markets, a 40 to 70 percent glide path on $3 million ended $3.3 million ahead of a fixed 60/40.

Is a market drop in my first year of retirement a reason to convert to a Roth?

Usually yes. A conversion is taxed on the value of the shares on the day they move, so converting after a 25 percent decline moves the same shares at 25 percent less tax, and the recovery then happens inside the Roth. Conversions made since 2018 cannot be reversed, so size the amount to the top of your target bracket and check the Medicare IRMAA lookback before you convert.

When does sequence of returns risk stop mattering?

It fades after roughly the first ten years, once the portfolio has grown relative to the remaining withdrawals, which is why William Bengen says plans that fail do so in the first ten to twelve years. For a woman retiring at 60 who may live to 95, the window is the same length but the horizon after it is longer, so the plan has to survive the first decade with enough left to fund twenty-five more years.

How ready is your plan for this?

The withdrawal rate, the guardrails, the allocation on your retirement date, and the reserve are four decisions, and they only work together. My Retirement Readiness Assessment takes about ten minutes and shows you where your plan is solid and where it is exposed, including how it would have held up in 2000 or 2008.

---

Hazel Secco, CFP®, CDFA®, is the founder of Align Financial Solutions, a fee-only fiduciary wealth management firm for high-net-worth women with complex financial lives: retirement, equity compensation, tax, and estate as one coordinated plan.

Book a free 15-minute Align Call: https://alignfinancialsolutions.com/book-a-call/
Whether we work together or not, you’ll walk away with clarity on your best next step.

🌐 https://alignfinancialsolutions.com
📺 https://www.youtube.com/@AlignYourRetirement
💼 https://linkedin.com/in/hazel-secco

Listen to Align Your Retirement: Apple Podcasts · Spotify · YouTube

Advisory services are offered through Align Financial Solutions LLC, an investment advisor in the State of New Jersey. This article is for educational purposes only and is not personalized tax, legal, or investment advice. The historical replay uses published S&P 500 and Bloomberg U.S. Aggregate index total returns for 2000 through 2025 and assumed returns of 7 percent for stocks and 4 percent for bonds for 2026 through 2029; it excludes taxes, fees, and expenses, and past performance does not guarantee future results. The retiree described is a hypothetical illustration, not a client. Tax figures are for 2026 and subject to change. Consult a qualified professional about your own situation.

Sources

1. Slickcharts, S&P 500 total returns by year: https://www.slickcharts.com/sp500/returns
2. NYU Stern (Damodaran), Historical returns on stocks, bonds and bills: https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/histretSP.html
3. Bloomberg, Looking back at 2024: fixed income (Bloomberg U.S. Aggregate 2024 return): https://www.bloomberg.com/professional/insights/markets/looking-back-at-2024-fixed-income/
4. YCharts, Bloomberg U.S. Aggregate Bond Index total return: https://ycharts.com/indices/%5EBBUSATR
5. LCG Associates, Bloomberg Aggregate then and now (calendar-year returns): https://www.lcgassociates.com/wp-content/uploads/2024/03/Market-Perspectives-Bloomberg-Aggregate-Then-Now-Final.pdf
6. Morningstar, The State of Retirement Income: 2025 (3.9% base rate; 5.2% with guardrails): https://www.morningstar.com/retirement/whats-safe-retirement-withdrawal-rate-2026
7. Bengen, W. P., Determining Withdrawal Rates Using Historical Data, Journal of Financial Planning, October 1994: https://www.financialplanningassociation.org/sites/default/files/2021-04/MAR04%20Determining%20Withdrawal%20Rates%20Using%20Historical%20Data.pdf
8. Advisor Perspectives, Bill Bengen boosts the 4% rule to 4.7% (2025): https://www.advisorperspectives.com/articles/2025/08/29/bill-bengen-boosts-the-4-rule-to-4-7
9. Pfau, W. D. and Kitces, M. E., Reducing Retirement Risk with a Rising Equity Glide Path, Journal of Financial Planning, January 2014: https://www.financialplanningassociation.org/article/journal/JAN14-reducing-retirement-risk-rising-equity-glide-path
10. Guyton, J. T. and Klinger, W. J., Decision Rules and Maximum Initial Withdrawal Rates, Journal of Financial Planning, March 2006: https://www.financialplanningassociation.org/sites/default/files/2021-11/2006%20-%20Guyton%20and%20Klinger%20-%20Decision%20Rules%20and%20SWR%20(1).PDF
11. Kitces, M., Managing the portfolio size effect with a bond tent in the retirement red zone: https://www.kitces.com/blog/managing-portfolio-size-effect-with-bond-tent-in-retirement-red-zone/
12. Kitces, M., Managing sequence of return risk with bucket strategies vs a total return rebalancing approach: https://www.kitces.com/blog/managing-sequence-of-return-risk-with-bucket-strategies-vs-a-total-return-rebalancing-approach/
13. Kitces, M., Guyton-Klinger guardrails and risk-based guardrails: https://www.kitces.com/blog/guyton-klinger-guardrails-retirement-income-rules-risk-based/
14. Social Security Administration, Period life table, 2023 (2026 Trustees Report): https://www.ssa.gov/oact/STATS/table4c6.html
15. 26 CFR § 1.408A-4, Converting amounts to Roth IRAs (amount includible in income): https://www.law.cornell.edu/cfr/text/26/1.408A-4
16. IRS Publication 590-A, no recharacterization of conversions made in 2018 or later: https://www.irs.gov/publications/p590a
17. IRS, Tax inflation adjustments for tax year 2026 (Rev. Proc. 2025-32): https://www.irs.gov/newsroom/irs-releases-tax-inflation-adjustments-for-tax-year-2026-including-amendments-from-the-one-big-beautiful-bill
18. New Jersey Division of Taxation, Capital gains: https://www.nj.gov/treasury/taxation/njit9.shtml
19. New Jersey Division of Taxation, 2025 NJ-1040 instructions (no loss carryover; no netting across income categories): https://www.nj.gov/treasury/taxation/pdf/current/1040i.pdf
20. Bogleheads forum threads: Retiring Soon and Worried About Sequence of Returns Risk https://www.bogleheads.org/forum/viewtopic.php?t=437253 ; Retiring into a recession https://www.bogleheads.org/forum/viewtopic.php?t=412284 ; When Does Sequence Risk Kick In? https://www.bogleheads.org/forum/viewtopic.php?t=458227