- Sequence of Returns Risk and Retirement Withdrawal Guardrails
- Longevity Risk: Why Pooling Makes Insurance Cheaper (you are here)
This is part 2 of 2 of our guide to Retirement Risks. It picks up where Sequence of Returns Risk and Retirement Withdrawal Guardrails leaves off, and it is written to stand on its own: the key ideas are restated where you need them.
A 65-year-old man’s period life expectancy is 18.12 years and a woman’s 20.66, but an individual cannot plan for the average: about 8.0% of 65-year-old men (6,333 ÷ 79,084) and 14.4% of women (12,612 ÷ 87,399) reach 95. Self-insuring means funding the tail. Ignoring investment returns to isolate the effect, $1,000,000 spread over 30 years (to 95) supports $1,000,000 ÷ 30 = $33,333 a year; spread over a man’s 18.12-year expectancy it would support $1,000,000 ÷ 18.12 = $55,188. A large pool can pay close to the second figure because members who die early leave their remaining money to those who live long; insurers call these mortality credits. That is the economic case for annuities and for delaying Social Security (10.3, 10.10), and why both become more valuable the longer you are likely to live. The cost is liquidity and, for a private annuity, insurer and inflation risk.
Guardrails turn the fixed 4% rule into a feedback loop. The best-known version, from planners Jonathan Guyton and William Klinger, uses three rules, simplified here:
- Upper guardrail — if this year’s planned withdrawal, measured as a percentage of the current portfolio, is more than 20% above your starting rate, cut the dollar withdrawal by 10%.
- Lower guardrail — if that percentage is more than 20% below the starting rate, because the portfolio has grown, raise the withdrawal by 10%.
- Skip a raise — after a year with a negative return, don’t add the inflation increase.
Example: start at $50,000 on $1,000,000, a 5% rate, so the guardrails sit at 5% × 1.2 = 6% and 5% × 0.8 = 4%. Next year’s planned withdrawal is $50,000 × 1.03 = $51,500. If the portfolio has fallen to $800,000, that is $51,500 ÷ $800,000 = 6.4%, above 6%: spending drops by 10% to $46,350. If it has grown to $1,350,000, the rate is $51,500 ÷ $1,350,000 = 3.8%, below 4%: spending rises 10% to $56,650. (In Guyton and Klinger’s 2006 paper the cutting rule is switched off in the final 15 years of the planned horizon; the version here keeps it on, which is more cautious.) Guardrails allow a higher starting rate because they accept that income will vary.
The bucket approach organizes the same portfolio by time rather than by rule: roughly one to two years of spending in cash, the next several years in bonds, and the rest in stocks (Part 7.4). Withdrawals come from cash; good years refill it. The total risk is unchanged, but knowing the next two years are already set aside makes it easier not to sell stocks in a crash.
Floor and upside separates needs from wants. Essential spending — housing, food, insurance, healthcare — is covered by income that cannot fall with markets: Social Security, a pension, an annuity (10.10), or a ladder of Treasury bonds. Discretionary spending such as travel and gifts comes from the invested portfolio and flexes with it. A household whose floor covers its essentials can live with variable withdrawals far more comfortably than one whose groceries depend on the market.
Each plan starts with $1,000,000 and meets Retiree A’s fifteen returns from the worked example above, then an assumed 6% a year for years 16–30. Inflation is 3%. Withdrawals come out at the start of each year; figures are nominal and before tax and fees.
| Measure | Fixed 4%, raised for inflation | Constant 4% of each year’s balance | Guardrails, 5% start (rules above) |
|---|---|---|---|
| First-year withdrawal | $40,000 | $40,000 | $50,000 |
| Lowest withdrawal | $40,000 (year 1) | $27,232 (year 4) | $35,847 (year 6) |
| Withdrawn, years 1–15 | $743,957 | $512,464 | $622,658 |
| Balance after 15 years | $741,927 | $1,233,766 | $965,051 |
| Year 30 | Cannot pay year 30’s $94,263 | $1,602,837 left | $982,876 left |
| Withdrawn, 30 years | $1,808,754 over 29 years | $1,351,261 | $1,440,870 |
The guardrail plan cut four times in its first six years: $50,000 → $45,000 → $40,500 → $37,544, then an inflation raise to $38,670 in year 5, then $35,847 in year 6; each cut was triggered because the planned withdrawal exceeded 6% of the portfolio. In today’s dollars its lowest year was $35,847 ÷ 1.035 = $30,922, a 38% real cut, so guardrails work only for a household that can absorb that. The constant-percentage plan can never run out and its final balance is the same in either order of returns, because a percentage withdrawal multiplies rather than subtracts; the risk moves entirely into income, which fell to $27,232 by year 4.
The flip point. On A’s order, the highest fixed inflation-adjusted starting rate that lasts all 30 years is 3.95% of the starting balance, found by trying rates until the money lasts exactly 30 years. On B’s order (the same fifteen returns reversed) it is 5.45%, and with a smooth 6% every year it is 4.90%. A fixed 4% plan survives or fails according to an order of returns you cannot know in advance.
Sequence risk is like driving across a desert with a fuel leak that is worst at the start. Lose half the tank in the first hour and no amount of fuel-efficient driving later gets you across; lose the same amount near the end, and you coast in. Guardrails are slowing down as soon as you see the gauge drop.
The tools above assume a 30-year retirement, a diversified portfolio and spending that can bend. These situations change the answer:
| Situation | What changes | Why | Number or rule |
|---|---|---|---|
| The market falls just before your retirement date | Your planned withdrawal starts at a higher rate | The same dollars are a larger share of a smaller portfolio | $40,000 on $1,000,000 is 4%; on $800,000, 5%: set the starting rate on the actual balance |
| You retire at 50 or earlier | A 45-year horizon to 95 | More years for a bad decade to occur and for inflation to compound | Use a fixed rate below 10.2’s 30-year figures, or flexible rules; plan for the 59½ penalty rules (edge cases in 10.5) |
| Social Security and a pension cover all essentials | Guardrail cuts touch only discretionary spending | The floor removes the cost of cutting | A higher starting rate is more defensible |
| A large one-off cost (roof, car, care) | The year’s withdrawal rate jumps and may trip a guardrail | The rule sees a percentage, not a reason | Fund known lumpy costs from a separate reserve in cash or short Treasuries |
| A spouse much younger than you | The horizon is set by the younger spouse | The plan must last until the second death | Plan to the younger spouse’s 95 |
| Inflation spikes early | Social Security keeps up only after a lag | The COLA compares CPI-W third-quarter averages and is paid from January | Fixed pensions and annuities do not adjust at all (10.4, 10.10) |
| RMDs exceed what you spend | The excess must still leave the IRA | RMDs cannot be rolled back in | Reinvest it in a taxable account; it is not a signal to spend more |
None of these risks can be eliminated, but each has a known counter: flexible spending for sequence risk, inflation-adjusted income for inflation risk, planning to 95 rather than 85 for longevity risk, and a guaranteed floor so that a bad decade costs comfort rather than necessities.
Each January, divide this year’s planned withdrawal (last year’s amount plus inflation, or unchanged after a negative-return year) by the current portfolio. If the result is more than 1.2 times your starting rate, then cut the dollar withdrawal by 10%; if it is less than 0.8 times, raise it by 10%; otherwise take the planned amount. It assumes a diversified portfolio with a meaningful stock share, a horizon of about 30 years, and spending that can fall by more than a third in real terms for several years, as it did in the worked example above. Ignore it when your essential spending is not covered by Social Security, a pension or other guaranteed income: then a cut lands on necessities, and the first step is to build that floor (10.10) and use a lower fixed rate for the rest.
Holding withdrawals fixed through a bad first decade. On Retiree A’s returns, the fixed plan withdrew $743,957 − $622,658 = $121,299 more than the guardrail plan over fifteen years and ended those years with $965,051 − $741,927 = $223,124 less. Followed for thirty years, the fixed plan could not pay year 30’s withdrawal; the guardrail plan still held $982,876. The extra spending came early, when it felt affordable, and the cost arrived in the retiree’s 90s, when nothing could be done about it. How to avoid it: decide your adjustment rule before you retire, write down the starting rate, and check it every January, most closely in the first ten years.
What is sequence of returns risk?
It is the risk that poor returns arrive early in retirement, while you are withdrawing. Losses force sales at low prices, and the shares sold miss the recovery, so two retirees with the same average return can end decades apart. In this chapter’s example, the same fifteen returns in opposite order left balances $635,129 apart.
How long will $1 million last in retirement?
At $40,000 a year rising 3% for inflation, about 28 years at a steady 4% return, 33 years at 5% and 42 years at 6%. Real returns are not steady: on Retiree A’s poor early sequence, followed by 6% a year, the money paid 29 full years even though the 15-year average return was also 6%.
What is the guardrails withdrawal strategy?
A rule that adjusts spending when your withdrawal rate drifts. In the Guyton-Klinger version, you cut 10% if the current rate is more than 20% above your starting rate, raise 10% if it is more than 20% below, and skip the inflation raise after a losing year. It permits a higher starting rate in exchange for variable income.
How do I protect my retirement savings from a market crash?
You cannot avoid crashes, but you can avoid selling into them. Hold one to two years of spending in cash, a further cushion in bonds, cover essentials with guaranteed income, and reduce withdrawals by rule when the portfolio falls. Selling stocks after a crash turns a temporary loss into a permanent one.
What age should I plan to live to for retirement?
Plan to at least 95. Under SSA’s 2023 period life table, 8.0% of 65-year-old men and 14.4% of 65-year-old women reach 95, and for a couple the chance that at least one does is about 21%. Period tables tend to understate how long today’s retirees will live.
Once you are withdrawing, the order of returns matters: losses early in retirement force sales at low prices, so two retirees with identical returns in opposite order were $635,129 apart after fifteen years, and sequence-of-returns risk is concentrated in the first five to ten years. Inflation at 3% halves a fixed income’s real value in about 24 years, and a 65-year-old couple has about a 51% chance that at least one partner reaches 90, so plan to 95 rather than 85. Guardrails turn the 4% rule into a feedback loop, cutting the withdrawal 10% when its rate rises more than 20% above the starting rate and raising it 10% when it falls more than 20% below, while buckets make it easier to hold stocks through a crash. Cover essentials with a guaranteed floor of income that cannot fall with markets, and let discretionary spending flex with the portfolio.
Four questions on this chapter. Decide on your answer first, then click “Reveal Answer.”
1. A retiree began with a $60,000 withdrawal from $1,200,000 (a 5% rate) and follows Guyton-Klinger guardrails. After a year of positive returns the portfolio is $1,650,000. With 3% inflation, what should next year’s withdrawal be?
- $61,800
- $67,980
- $66,000
- $82,500
Reveal Answer
Answer: B. Planned $60,000 × 1.03 = $61,800 is 3.7% of $1,650,000, below the 4% lower guardrail (5% × 0.8), so raise it 10%: $67,980. $61,800 ignores the guardrail, $66,000 raises last year’s amount, $82,500 resets to 5% of the new balance. (Part 10.7)
2. A retiree’s essential spending is $45,000 a year and Social Security pays $33,000. Under a floor-and-upside plan, what should cover the remaining $12,000 of essentials?
- A higher stock allocation so the portfolio outgrows the gap
- Portfolio withdrawals adjusted each year by guardrails
- An income annuity or Treasury ladder paying $12,000 a year
- Two years of spending kept in a cash bucket and refilled
Reveal Answer
Answer: C. Floor and upside covers essentials with income that cannot fall with markets: Social Security, a pension, an annuity or a Treasury ladder. Guardrails, buckets and stock growth still depend on the portfolio, so they serve discretionary spending. (Part 10.7)
3. Which spending rule leaves the same ending balance whatever the order in which a given set of annual returns arrives?
- Guardrails that start at 5% and adjust by 10%
- Withdrawing $40,000 raised 3% a year for inflation
- Withdrawing a constant 4% of each year’s balance
- Withdrawing a fixed $40,000 every year
Reveal Answer
Answer: C. A percentage withdrawal multiplies the balance by (1 − 4%) each year, and multiplication does not depend on order, so the risk moves into income instead. Dollar withdrawals subtract, so early losses do lasting damage. (Part 10.7)
4. Worked problem: If spending of $50,000 rises 3% a year, what is it after 25 years?
Reveal Answer
Answer: $50,000 × 1.0325 = $104,689.
- Actuarial life table — Survival odds
- SSA: LatestCOLA
- IRS RMD FAQs