Income Planning · 11 min read
Retirement Income and Withdrawal Rates: From One Number to a Coordinated Plan
Current research places a sustainable starting withdrawal rate between about 3.9% and 4.7% of a diversified portfolio, so a 4% withdrawal falls inside the range. The rate matters less than the structure around it: a two-to-three-year cash reserve, predictable bond income, pre-agreed spending guardrails, and an account order coordinated with taxes and Social Security.
Most retirement conversations eventually arrive at one percentage. Often it is "4%," and that figure quietly becomes the plan. We treat a starting rate as a useful anchor that answers a narrower question than most families think they are asking. The durable answer comes from the structure around it: where the cash sits, which account funds each dollar, when Social Security begins, and what the household will do when markets disappoint.
What the Withdrawal-Rate Question Is Really Asking
"How much can we spend?" sounds like a single question. In practice it bundles at least four separate ones, and each carries its own evidence:
- The sustainability question. Across a long horizon and many possible market paths, what first-year withdrawal has held up in most of them?
- The timing question. What happens if the worst returns arrive in the first few years, while withdrawals are already underway?
- The liquidity question. Where does next year's spending come from if stocks are down when the bills are due?
- The coordination question. Which account funds each withdrawal, and what does that choice do to taxes, Medicare premiums and Social Security?
A rule of thumb speaks to the first question only. Households that struggle rarely chose a reckless percentage; more often, they had no answer ready for the other three.
The Evidence: Two Research Programs, One Workable Range
The familiar figure traces back to 1994, when William Bengen tested withdrawals against U.S. market history reaching back to 1926. Take 4% of the starting portfolio in year one, then raise that dollar amount each year for inflation: under those rules, no 30-year stretch in the record exhausted the portfolio, a result his updated framework still reaches.
Bengen revisited the work in August 2025. With small-cap stocks and a broader mix of asset classes added to the test portfolio, the worst-case starting rate he calls "Safemax" rose to 4.7%. Averaged over every historical starting year, the figure was closer to 7.1%. That gap is a reminder that the headline number is built from the unluckiest retirement cohorts, not the typical one. (Advisor Perspectives, August 2025)
Morningstar approaches the problem from the other direction. Its annual study projects forward-looking returns rather than replaying history, and it targets a 90% chance that the portfolio lasts 30 years. For someone retiring in 2026 with a portfolio holding 30% to 50% in equities, the December 2025 edition puts the starting rate at 3.9%, up from 3.7% in the prior year's study. (Morningstar, December 2025; FA Magazine)
Three qualifiers travel with any of these figures:
- Probability is not certainty. A 90% success rate still leaves roughly one simulated path in ten that runs short.
- The horizon is 30 years. A household retiring in its 50s may be planning for 35 to 40 years, and the math tightens accordingly.
- Diversification is assumed. Portfolios held mostly in cash or very conservative holdings can fall short of what the models describe.
Flexibility moves the range upward. Households that accept modest spending adjustments tied to portfolio results, usually described as guardrails, may be able to begin nearer 4.7% to 5.7% without a material rise in depletion risk. Morningstar's work finds that pairing that flexibility with a later Social Security claim can support starting rates approaching 5.7%.
Why the order of returns matters as much as the average
Two retirees can post the same average return over three decades and finish far apart because the bad years came first for one of them. Spending from a falling portfolio means selling more shares at depressed prices, and those shares are gone before any rebound.
Morningstar research reported by CNBC in March 2025 put numbers to the effect. Where a portfolio lost at least 15% in the first year of retirement while the retiree withdrew 3.3%, the odds of running out of money within 30 years were six times higher than for a retiree whose first year was positive. (CNBC, March 2025) The historical worst cases behind the original 4% figure, including the 1966 and 1973 retirement cohorts, were largely stories of weak early returns. For that reason we manage sequence risk directly rather than hoping the percentage alone absorbs it. Our guide to sequence risk goes further into the mechanics.
The Structure Around the Rate
Three tools do most of the work of keeping a withdrawal rate intact through a bad stretch, and each is most useful alongside the others.
A time-segmented reserve
Often called a bucket approach, this divides the portfolio by when each dollar is expected to be spent. The point is mechanical: near-term spending should never depend on selling stocks during a decline.
| Segment | Spending window | Typical holdings | Role in the plan |
|---|---|---|---|
| Near-term reserve | Years 1–3 | Cash, money market funds, short-term Treasuries | Pays current living costs; untouched by equity swings |
| Refill layer | Years 4–10 | Intermediate bonds, bond funds, TIPS | Replenishes the reserve; absorbs a prolonged downturn |
| Growth layer | Year 10 and beyond | Equities, real assets | Long-run growth; sold down after markets recover |
With two to three years of spending in the reserve, a 20% market decline does not require a single equity sale. The household draws from cash and lets the growth layer wait. The discipline is in the refill direction: gains move from growth to refill to reserve, never stocks sold at the bottom to top up cash. Handled consistently, a stressful decision becomes a routine one.
A bond ladder for known cash flows
A ladder holds individual bonds or Treasuries maturing in successive years. Unlike a bond fund, each rung returns a known amount on a known date, whatever prices do in between, which makes it a natural backbone for the reserve and refill layers.
Yields in mid-2026 give ladders more appeal than they have had in over a decade. Approximate Treasury levels at the time of writing:
- 5-year: about 4.2%
- 10-year: about 4.5%
- 30-year: about 5.0%
- 10-year TIPS: a real yield of about 2.1% above inflation
(U.S. Department of the Treasury, Daily Treasury Yield Curve Rates)
Inflation is the risk a nominal ladder does not address. Blending TIPS into 20% to 30% of the ladder ties part of the income to consumer prices. Bengen himself identified inflation as the leading long-term threat to retirement income in September 2025 commentary. (CNBC, September 2025)
Guardrails for spending
A guardrail policy sets the rules for adjustment in advance. When the portfolio runs meaningfully ahead of its projected path, spending rises modestly. When it falls meaningfully behind, spending is trimmed modestly. Even small, pre-agreed adjustments of this kind can raise the sustainable starting rate by roughly 0.5 to 1.5 percentage points compared with a withdrawal that never flexes.
A Hypothetical Household, Worked Through
This illustration is hypothetical and simplified. It is not based on any client, does not project investment returns, and is not a recommendation for any household.
Consider a couple entering retirement with a $1,500,000 diversified portfolio. Their planned spending is $100,000 a year, and Social Security will provide $40,000 of it. The portfolio has to cover the other $60,000.
- Implied withdrawal rate: $60,000 ÷ $1,500,000 = 4.0%.
- At Morningstar's 3.9%: $1,500,000 × 3.9% = $58,500, or $1,500 a year below their target.
- At Bengen's 4.7%: $1,500,000 × 4.7% = $70,500, or $10,500 a year of room above it.
On the percentage alone, their plan sits inside the research range. The structure is where the real decisions appear.
The reserve. Two to three years of portfolio withdrawals is $60,000 × 2 = $120,000 up to $60,000 × 3 = $180,000, or 8% to 12% of the portfolio held in cash and short-term Treasuries.
A full ladder. Ten rungs of $60,000 face value, maturing in years one through ten, total $600,000. That is 40% of the portfolio committed to the ladder, which makes the ladder an asset-allocation decision as much as an income one.
A poor first year. Suppose the portfolio falls 15% in year one and the $60,000 withdrawal is then taken. The balance goes from $1,500,000 to $1,275,000 ($1,500,000 × 0.85), then to $1,215,000 after the withdrawal. Next year, even before any inflation increase, $60,000 equals about 4.9% of the balance ($60,000 ÷ $1,215,000). This is the moment the reserve and guardrails exist for: spending comes from cash, and any pre-agreed trim is already understood.
How We Approach This in a Plan
We set a withdrawal range, not a single number, and test it against the decisions that sit next to it. That coordination is where most of the value in an income plan lives.
- Taxes. A $60,000 withdrawal can be fully taxable, partly taxable or close to tax-free depending on whether it comes from a pre-tax IRA, a taxable account or a Roth. See our guide to withdrawal sequencing.
- Social Security timing. A later claim means heavier withdrawals in the bridge years and lighter ones afterward; those bridge years concentrate sequence risk and need their own reserve. See our Social Security claiming framework.
- Medicare and IRMAA. Withdrawals from pre-tax accounts raise income, and Medicare premium surcharges are set from the tax return filed two years earlier. (SSA, Medicare premiums) A large one-time withdrawal can surface later as a higher premium; see the IRMAA cliff.
- Required distributions. Once required minimum distributions begin, part of the withdrawal from pre-tax accounts is set by the IRS rather than by the household's budget. (IRS RMD FAQs) We plan the earlier years with that later floor in mind.
- Account location. Bond interest is taxed as ordinary income, so the account holding the ladder and reserve shapes the after-tax result of the whole structure.
- Sequence risk. The reserve size, the ladder length and the guardrail triggers are set together, so a weak first few years has a written response before it happens.
The output is a decision architecture the household can follow in any market: which account to draw from, when to refill the reserve, and when spending adjusts. We revisit it as balances, tax law and health change.
Questions Worth Answering Before You Decide
Before settling on a figure, we work through a short checklist:
- What share of annual spending is covered by Social Security, pensions or other income, and what share has to come from the portfolio?
- How long should the plan run? Is 30 years enough, or does an earlier retirement or family longevity argue for 35 to 40?
- How much spending flexibility is realistic, and which expenses could be trimmed in a poor year without real hardship?
- Is there a cash reserve covering two to three years of withdrawals, and where does it sit?
- In what order will taxable, pre-tax and Roth accounts be drawn, and how does that order affect each year's tax bracket?
- When will each spouse claim Social Security, and how will the years before claiming be funded?
- How close does projected income sit to the IRMAA thresholds two years ahead?
- What portfolio level would trigger a spending cut, and what level would allow a raise?
What This Does Not Mean
A research-backed range is not a spending entitlement. A 3.9% or 4.7% starting rate describes how portfolios behaved in a model or in history; it is not a promise about the next 30 years. Nor is a lower rate automatically safer: underspending has its own cost in goals left unmet.
Buckets and ladders do not remove market risk; they change when and how it is felt. A bond ladder fixes nominal cash flows, and without TIPS those dollars can lose purchasing power. Guardrails only help if the household follows them in a difficult year. Nothing here is a recommendation to buy or sell any security or to adopt a specific withdrawal rate.
Frequently Asked Questions
Q: Does the 4% rule still hold up in 2026?
As a starting reference, yes. Morningstar puts the 2026 starting rate at 3.9% (30% to 50% equities, 90% confidence, 30 years), and Bengen's 2025 update puts the historical worst case at 4.7% with a broader asset mix. A fixed 4% sits between the two for a diversified, roughly 30-year plan.
Q: Why do the first years of retirement carry so much weight?
Losses taken while withdrawals continue force more shares to be sold at low prices, and those shares cannot recover. Morningstar research reported in March 2025 found that a first-year loss of at least 15%, with a 3.3% withdrawal, made depletion within 30 years six times more likely than a positive first year.
Q: Do bond ladders make sense at today's yields?
With 10-year Treasuries near 4.5% and 30-year yields near 5.0% as of mid-2026, ladders are more attractive than they have been in over a decade. Bonds held to maturity deliver known cash flows on known dates; adding TIPS brings some inflation protection.
Q: What are spending guardrails?
Guardrails are pre-agreed rules that raise spending modestly when the portfolio runs well ahead of its projected path and trim it modestly when it falls behind. Even small adjustments can lift the sustainable starting rate by about 0.5 to 1.5 percentage points compared with a rigid withdrawal.
Q: How does the withdrawal rate interact with taxes and Medicare?
The same withdrawal can carry very different tax costs depending on which account funds it, and pre-tax withdrawals raise the income that sets Medicare surcharges two years later. We set the rate together with the account order, the Social Security claiming age and the IRMAA thresholds.
Bring the decision in front of you.
The first conversation is 30 minutes. A recent statement helps; no preparation is needed.
Sources and further reading
Next steps
Keep reading
Get new guides by email.
An occasional email when we publish a new guide: the short answer, and what it may mean for your portfolio and your plan.
The views and opinions expressed here are those of The Financial Sciences Company as of the publish date and are provided for informational and educational purposes only. They are not personalized investment, tax, or legal advice. The Financial Sciences Company, LLC is a registered investment adviser. Registration does not imply a certain level of skill or training. Additional information is available in our Form ADV at adviserinfo.sec.gov.
General educational information, current as of 2026. Figures and rules change. For guidance specific to your situation, speak with a qualified professional.