Key Takeaways
Direct answer: A withdrawal-rate framework is the rule set a retiree uses to decide how much to spend from an investment portfolio each year. Fixed real withdrawals take a percentage of the starting portfolio value once, then increase that dollar amount annually for inflation regardless of performance. Percentage-of-portfolio withdrawals recalculate the percentage against the current balance every year, so income moves with the market. Guardrails strategies start like a fixed withdrawal but add rules that cut or raise spending after the withdrawal rate drifts too far from its starting point. Floor-and-upside spending separates essential needs, funded by reliable income, from discretionary spending, funded by the flexible portfolio. No single number or framework is universally "safe"; each makes a different trade-off between income stability and adaptability.
- The "4% rule" is one specific, historically researched framework (fixed real withdrawals), not a universal law of retirement math.
- Bengen's 1994 research and the 1998 Trinity Study are the two foundational historical studies behind that figure, both using U.S. stock and bond data and specific stock/bond allocation ranges.
- Percentage-of-portfolio withdrawals mathematically can never fully deplete an account, but their income can shrink to an inadequate level after a sustained downturn.
- Guardrails research found that dynamically cutting or raising spending in response to performance can support a meaningfully higher starting withdrawal rate than a fixed approach.
- Floor-and-upside spending is the only framework of the four that treats essential and discretionary spending as fundamentally different problems, rather than applying one rule to the whole portfolio.
- A withdrawal-rate framework is a voluntary spending choice; a required minimum distribution is a separate, mandatory IRS rule that can conflict with it in a given year.
Why There Is No Single "Safe" Number
Every published withdrawal-rate figure, 4%, 3.5%, 5%, or any other, is the output of a specific model built on specific assumptions: which historical period was tested, which stock and bond indexes represented the portfolio, what stock/bond allocation was assumed, how long the payout period needs to last, and whether the withdrawal amount responds to performance or stays fixed. Change any one of those assumptions and the resulting "safe" number changes with it. FINRA's own investor education on managing a retirement portfolio describes expert opinion as clustering in a 3% to 5% range for annual withdrawals, explicitly framing it as a range from varying expert judgment rather than a single agreed figure.
This guide is deliberately organized around frameworks, the rule sets that determine how a withdrawal amount is set and adjusted, rather than around any single percentage. Understanding which framework is actually being used, and what assumptions produced any specific number attached to it, matters more than memorizing a headline figure.
The Historical Research Behind the 4% Rule
Two pieces of research, published a few years apart, are the origin of the "4% rule" as it is commonly discussed today.
Financial planner William Bengen's October 1994 paper, "Determining Withdrawal Rates Using Historical Data," tested how long a portfolio would last under different fixed, inflation-adjusted withdrawal rates, using historical U.S. stock and bond returns for retirees starting at different points from 1926 onward. He found that a 4% initial withdrawal rate, increased annually for inflation, did not exhaust a portfolio holding 50% to 75% in stocks within 30 years in any historical period he tested, and recommended leaning toward the higher end of that stock range rather than the lower end.
A separate study published in the AAII Journal in February 1998 by three Trinity University finance professors, Philip Cooley, Carl Hubbard, and Daniel Walz, took a broader approach. Rather than testing one withdrawal rate, they calculated a "portfolio success rate", the percentage of all overlapping historical periods from 1926 to 1995 in which a given withdrawal rate and stock/bond allocation did not exhaust the portfolio, across withdrawal rates from 3% to 12%, payout periods from 15 to 30 years, and allocations from 100% stocks to 100% bonds. Their table of inflation-adjusted results (reproduced in part below) shows why 3% to 4% became the historically conservative benchmark, and why bond-heavy portfolios did comparatively worse at higher withdrawal rates despite feeling "safer."
| Allocation | Withdrawal Rate | 30-Year Success Rate (Inflation-Adjusted, 1926-1995) |
|---|---|---|
| 100% stocks | 4% | 95% |
| 75% stocks / 25% bonds | 4% | 98% |
| 50% stocks / 50% bonds | 4% | 95% |
| 75% stocks / 25% bonds | 6% | 68% |
| 50% stocks / 50% bonds | 6% | 51% |
| 100% bonds | 4% | 20% |
Two things stand out in this data. First, at a 4% withdrawal rate, every meaningfully stock-containing allocation the study tested succeeded in at least 95% of historical 30-year periods, once withdrawals were adjusted for inflation. Second, the all-bond portfolio, often assumed to be the "safe" choice, actually had the lowest success rate of the group shown here, because bonds offered little of the upside needed to outpace a fixed, inflation-adjusted withdrawal stream over three decades. The study's authors drew a specific conclusion from this pattern: because stocks supply the growth that keeps a withdrawal stream ahead of inflation over multi-decade periods, an all-bond or bond-heavy portfolio is a poor way to pursue safety at this task, and a meaningful equity allocation, at least half the portfolio in their assessment, tends to serve most retirees better than a more conservative-looking mix.
Both studies explicitly did not adjust for taxes or transaction costs, and both are historical backtests, not guarantees about future market behavior. Later researchers, including Guyton and Klinger discussed below, built directly on this foundation rather than replacing it.
Framework 1: Fixed Real Withdrawals
A fixed real (or constant-dollar, inflation-adjusted) withdrawal sets a percentage of the portfolio's value exactly once, in the first year of retirement, converts that into a dollar amount, and then increases that dollar amount every subsequent year by the rate of inflation, regardless of how the portfolio performs. This is the framework both Bengen's research and the Trinity Study tested, and it is what most people mean when they refer to "the 4% rule."
Its main advantage is predictability: a retiree budgeting on this framework knows, in real (inflation-adjusted) terms, exactly what income to expect every year, independent of market conditions. Its main weakness is exactly that same rigidity: because the dollar amount never adjusts downward after a poor sequence of returns, it is the framework most exposed to the sequence-of-returns risk described in Swoopr's Sequence-of-Returns Risk guide. The historical success rates above assume this specific, fixed behavior; a retiree who instead cuts spending after a bad year is no longer using a pure fixed-real framework, but is moving toward the guardrails approach covered later in this guide.
Framework 2: Percentage-of-Portfolio Withdrawals
A percentage-of-portfolio withdrawal recalculates the withdrawal amount every year as a fixed percentage of the portfolio's current balance, rather than its original starting balance. If a retiree withdraws 5% annually under this framework and the portfolio falls from $1,000,000 to $800,000, the next year's withdrawal automatically falls from $50,000 to $40,000; if the portfolio instead grows to $1,200,000, the withdrawal rises to $60,000.
This framework has one powerful mathematical property: because the withdrawal is always a percentage of whatever remains, the portfolio can never be reduced to exactly zero by the withdrawals alone. That guarantee, however, is a narrower promise than it sounds. A portfolio can still shrink to a level where a fixed percentage of it produces an income too small to live on, which is a practical failure even though it is not a mathematical one. This framework also produces the most volatile year-to-year income of the four covered here, since spending moves directly with portfolio performance with no smoothing mechanism.
Framework 3: Guardrails (Dynamic, Rules-Based Withdrawals)
A guardrails strategy, most closely associated with financial planners Jonathan Guyton and William Klinger's published research, starts like a fixed real withdrawal but adds explicit rules for when to deviate from it. In the commonly cited version of their approach, the retiree tracks their current withdrawal rate (that year's planned dollar withdrawal divided by the portfolio's current value) against their initial withdrawal rate (the percentage used to set the very first year's withdrawal):
- Capital preservation rule (the upper guardrail): if the current withdrawal rate rises 20% or more above the initial rate, because the portfolio has fallen in value, that year's dollar withdrawal is cut by 10%.
- Prosperity rule (the lower guardrail): if the current withdrawal rate falls 20% or more below the initial rate, because the portfolio has grown, that year's dollar withdrawal is raised by 10%.
A simple hypothetical illustrates the mechanic. Suppose a retiree starts with a $1,000,000 portfolio and a 5% initial withdrawal rate, a $50,000 first-year withdrawal. The upper guardrail sits at a current withdrawal rate of 6% (5% × 1.20). If a market downturn later leaves the portfolio at $800,000 while the planned withdrawal is still $50,000, the current rate is $50,000 ÷ $800,000 = 6.25%, above the 6% guardrail, triggering a 10% cut to $45,000. Conversely, if strong markets instead grow the portfolio to $1,400,000, the current rate falls to $50,000 ÷ $1,400,000 ≈ 3.6%, below the 4% lower guardrail (5% × 0.80), triggering a 10% raise to $55,000. This is a hypothetical illustration built to show the mechanism, not a projection of any real portfolio.
Guyton and Klinger's own published research found that applying rules like these, tested against stock-heavy portfolios (their published results point to at least roughly 65% equities), could support a meaningfully higher initial withdrawal rate, in the range of roughly 5.2% to 5.6%, than the static 4% figure from the earlier fixed-withdrawal research, precisely because the rules respond to a bad sequence rather than ignoring it. That higher starting rate is not a free upgrade: it depends on the retiree actually following the cut when the upper guardrail is triggered, not merely hoping to.
Framework 4: Floor-and-Upside Spending
A floor-and-upside (sometimes called essential/discretionary) strategy takes a different approach entirely: instead of applying one rule to the whole portfolio, it splits spending into two categories with two different funding sources. The floor covers essential, non-negotiable spending, such as housing, food, insurance, and health care, and is funded by the most reliable income sources available: Social Security, a pension if one exists, and sometimes an annuity or a dedicated bond or TIPS ladder built to mature when specific future expenses come due. The upside covers discretionary spending, travel, gifts, and similar wants rather than needs, and is funded by the remaining investment portfolio using whichever of the other three frameworks the retiree prefers for that portion.
The appeal of this approach is that it directly targets the consequence that matters most: a market downturn under this framework reduces discretionary spending, not essential spending, because the floor was deliberately built to be independent of portfolio performance. The trade-off is that funding a reliable floor, particularly through an annuity or a bond ladder, typically means giving up some of the higher long-run growth potential a fully invested portfolio might otherwise offer for that same money, and the discretionary portion is still exposed to ordinary market and sequence risk.
Comparing the Four Frameworks
| Framework | How the Amount Is Set | Income Stability | Best Suited For |
|---|---|---|---|
| Fixed real | Percentage of the starting portfolio, then inflation-adjusted every year | Highest; fully predictable in real terms | Retirees who most value simple, predictable budgeting and can accept a conservative starting rate |
| Percentage-of-portfolio | Percentage of the current portfolio, recalculated every year | Lowest; moves directly with the market | Retirees who can tolerate variable income and most want to avoid ever fully depleting the account |
| Guardrails | Fixed real, with a defined percentage cut or raise once the current rate drifts far enough from the starting rate | Moderate; adjusts occasionally, not annually | Retirees comfortable with occasional, rules-based spending changes in exchange for a potentially higher starting rate |
| Floor-and-upside | Essentials funded by guaranteed income; discretionary spending set by another framework applied to the remaining portfolio | Essential spending very stable; discretionary spending varies | Retirees who most want a hard floor under essential needs and can accept variability in discretionary spending |
How Sequence-of-Returns Risk Interacts with Each Framework
These four frameworks are, among other things, four different answers to sequence-of-returns risk, the subject of Swoopr's companion Sequence-of-Returns Risk guide. A fixed real withdrawal is the most exposed of the four: because the dollar amount never adjusts, a bad early sequence does its full, uncushioned damage, which is exactly why the historical research behind it requires a conservative starting rate in the first place. A percentage-of-portfolio withdrawal is, in a sense, automatically self-correcting against sequence risk, since a falling balance directly and immediately reduces the withdrawal, but that same automatic reduction is what produces its volatile income. Guardrails strategies are explicitly designed to counter sequence risk: the capital preservation rule exists specifically to reduce spending after the kind of early loss that does the most permanent damage. Floor-and-upside spending sidesteps the question for essential spending entirely, by funding it from sources that are not exposed to sequence risk in the first place, while leaving the discretionary portion exposed to whichever underlying framework manages it.
How This Differs from Required Minimum Distributions
A withdrawal-rate framework and a required minimum distribution (RMD) are frequently confused but solve different problems. A withdrawal-rate framework is a voluntary spending policy: the retiree chooses it to decide how much of a portfolio to spend, based on sustainability and personal goals. An RMD is a mandatory minimum amount the IRS requires be withdrawn from certain tax-advantaged retirement accounts, such as traditional IRAs and 401(k)s, once the account owner reaches the applicable age, regardless of whether that money is actually needed for spending. Swoopr's Required Minimum Distributions guide is the canonical, kept-current source for the specific ages and calculation rules; this page does not restate them, since they change with legislation and inflation adjustments.
The two can genuinely conflict. A guardrails or percentage-of-portfolio framework might call for a reduced withdrawal after a market downturn in the very same year an RMD requires a larger one, since the RMD calculation is based on the account's prior-year-end balance and the owner's age, not on any withdrawal framework's own logic. In that situation, the RMD amount is a legal floor beneath which the framework's own recommendation cannot go, even though the money withdrawn to satisfy it does not have to be spent; it can be reinvested in a taxable account if it is not needed.
Common Mistakes and Misconceptions
- Treating "4%" as a permanent, universal law. It is a historical research finding under specific assumptions about the market period tested, the stock/bond allocation used, and the payout period assumed, not a guarantee that applies to every portfolio or every era.
- Assuming a higher published rate is a free upgrade. Guyton and Klinger's higher 5.2% to 5.6% figures explicitly depend on actually following the guardrail adjustment rules, including real spending cuts after a bad sequence; withdrawing at that rate without the discipline to cut is a different, untested strategy.
- Ignoring taxes and fees when comparing research-based figures. Both the Trinity Study and Bengen's original research explicitly excluded taxes and transaction costs; a real portfolio's sustainable rate needs to account for both.
- Confusing a withdrawal-rate framework with an RMD. They are different mechanisms with different purposes, as the previous section explains, and they can require conflicting actions in the same year.
- Assuming percentage-of-portfolio withdrawals can never meaningfully fail. True only in the narrow sense that the balance can never hit exactly zero; a severely depleted balance can still produce income too small to meet real needs.
- Applying research from one allocation to a very different one. The Trinity Study's own data shows a 100%-bond portfolio performing worse than stock-containing portfolios at the same withdrawal rate; assuming "safer-feeling" assets automatically produce a safer withdrawal rate is not supported by the historical data.
Frequently Asked Questions
What is the 4% rule and where did it come from?
The 4% rule traces to financial planner William Bengen's 1994 research, which found that withdrawing 4% of a portfolio's starting value, then increasing that dollar amount each year for inflation, would not have exhausted a portfolio with 50 to 75% in stocks within 30 years in any historical U.S. period he tested. A separate 1998 study by three Trinity University professors, using overlapping historical periods from 1926 to 1995, reached similar conclusions and became known as the Trinity Study. Both are historical research findings under specific assumptions, not a guaranteed formula.
Is 4% still considered a safe withdrawal rate?
There is no single current consensus number; researchers have proposed both higher and lower figures over time depending on the interest-rate and valuation environment assumed at the start of retirement, the exact stock/bond mix, the payout period, and whether the strategy adjusts spending dynamically. FINRA's own investor education describes expert opinion as clustering in a 3% to 5% range rather than converging on one figure. This guide focuses on the frameworks and the research behind them rather than promoting a single number.
What's the difference between a fixed withdrawal and a percentage-of-portfolio withdrawal?
A fixed (constant real) withdrawal takes a percentage of the portfolio's value only once, in the first year, then increases that dollar amount annually for inflation regardless of how the portfolio performs. A percentage-of-portfolio withdrawal recalculates the percentage against the current balance every year, so the dollar amount rises and falls with the portfolio. The fixed method gives more predictable income but carries a real historical risk of depletion at higher withdrawal rates; the percentage method can never mathematically reduce the portfolio to exactly zero, but its income can become uncomfortably small after a sustained downturn.
How do guardrails withdrawal rules work?
A guardrails strategy, based on research by Jonathan Guyton and William Klinger, starts like a fixed withdrawal but adds rules that adjust spending when the current withdrawal rate drifts too far from its starting point. In their published version, if the current withdrawal rate rises 20% above the initial rate due to poor performance, dollar withdrawals are cut by 10%; if it falls 20% below the initial rate due to strong performance, withdrawals are raised by 10%. Their research found this dynamic approach could support a higher initial withdrawal rate, roughly 5.2% to 5.6% for stock-heavy portfolios, than the static 4% figure, precisely because it responds to a bad sequence instead of ignoring it.
What is a floor-and-upside retirement spending strategy?
A floor-and-upside strategy splits retirement spending into two layers: an essential-spending floor funded by highly reliable income sources, such as Social Security, a pension, or an annuity, and a discretionary-spending upside funded by the remaining investment portfolio using one of the other withdrawal frameworks. The floor is designed to hold up regardless of market performance; the discretionary portion is allowed to flex with the portfolio's actual returns, since a shortfall there does not threaten essential needs.
Is a withdrawal-rate framework the same as a required minimum distribution?
No. A withdrawal-rate framework is a voluntary spending policy an investor chooses to decide how much of a portfolio to spend each year. A required minimum distribution is a mandatory minimum amount the IRS requires be withdrawn from certain tax-advantaged retirement accounts by rule, regardless of whether the investor actually wants or needs to spend it. The two can conflict: an RMD can force a larger distribution than a chosen withdrawal framework would otherwise call for in a given year.
Which withdrawal framework is best?
None of the four frameworks is universally best; each makes a different trade-off between income stability and adaptability to market performance. A fixed real withdrawal maximizes predictability at the cost of never adjusting to a bad sequence. A percentage-of-portfolio approach adapts automatically but creates variable income. Guardrails add rules-based flexibility to a largely fixed approach. Floor-and-upside protects essential spending specifically while allowing discretionary spending to vary. The right choice depends on an individual's risk tolerance, other income sources, and need for predictable spending, not a universal ranking.
How do taxes change the amount a withdrawal framework actually delivers?
A withdrawal rate describes gross withdrawals from the portfolio, and what reaches spending depends on which accounts the money comes from. A dollar from a traditional account is generally taxed as ordinary income on withdrawal, a dollar from a Roth account may not be taxed at all, and a dollar from a taxable account triggers tax only on the gain. Two retirees withdrawing the same percentage can therefore spend materially different amounts, which is why frameworks are usually paired with a withdrawal sequencing plan.
How do these frameworks handle spending that is not level through retirement?
Most are built around a smooth inflation-adjusted stream, which is a simplification. Actual spending patterns often decline in real terms through the middle of retirement and can rise later with healthcare costs, and one-time items such as a house move or a vehicle arrive as lumps. A framework can be adapted by modeling the intended spending path directly rather than a constant one, which changes the sustainable starting figure in either direction depending on the shape.
References
This guide is based on publicly available AAII Journal, Journal of Financial Planning, and FINRA materials, verified in August 2026. Key sources include:
- Philip L. Cooley, Carl M. Hubbard, and Daniel T. Walz, "Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable," AAII Journal, February 1998: the Trinity Study's full portfolio-success-rate tables, including the inflation-adjusted results summarized above.
- William P. Bengen, "Determining Withdrawal Rates Using Historical Data," Journal of Financial Planning, October 1994: the original 4% rule research.
- Jonathan T. Guyton and William J. Klinger, "Decision Rules and Maximum Initial Withdrawal Rates," Journal of Financial Planning, March 2006: the guardrails decision rules described above.
- FINRA: Managing Your Retirement Portfolio: current investor-education framing of the 3% to 5% expert-opinion range referenced above.
This content was reviewed by the Swoopr Editorial Team in August 2026 and reflects publicly available information at that time. The guardrails illustration in this guide is an original, hypothetical example built to demonstrate the published mechanism; it is not a projection of any real portfolio's performance. All withdrawal-rate figures cited are historical research findings, not personalized financial advice, guarantees, or recommendations for any individual's retirement plan.
Conclusion
"How much can I withdraw?" has no single correct answer because it depends on which rule set is being used to answer it. Fixed real withdrawals offer predictability at the cost of rigidity. Percentage-of-portfolio withdrawals adapt automatically at the cost of stable income. Guardrails add disciplined, rules-based flexibility to a largely fixed approach. Floor-and-upside spending protects essential needs specifically, independent of market performance. Understanding these as four distinct frameworks, each with real historical research behind it, is more useful than searching for one magic percentage that applies to every retiree.