In 1994, financial planner William Bengen published a study that would reshape retirement planning for a generation. Using rolling 30-year periods from 1926 to 1992, he found that a retiree withdrawing 4% of their initial portfolio annually (adjusted for inflation) would never have run out of money over any historical period. The "4% rule" was born, and it quickly became the default framework for retirement income planning.
Three decades later, the rule remains the most widely cited guideline in personal finance. But the research landscape has evolved substantially. International data, dynamic withdrawal strategies, and valuation-adjusted frameworks all suggest the original rule is simultaneously too conservative for some retirees and too aggressive for others. The difference depends on three variables: the country of retirement, the valuation environment at the start, and whether the retiree is willing to adjust withdrawals based on portfolio performance.
The Foundation: What Bengen and Trinity Actually Found
Bengen's original analysis examined a 50/50 stock/bond portfolio using US data from 1926 to 1992. He identified the maximum sustainable withdrawal rate (the highest initial withdrawal rate, adjusted annually for inflation, that would not exhaust the portfolio over 30 years) for each retirement cohort. The worst-case cohort was 1966, which retired into a period of rising inflation, stagnant equity returns, and negative real bond yields.
The Trinity Study (Cooley, Hubbard, and Walz, 1998) extended this analysis by testing multiple withdrawal rates and asset allocations, reporting success rates rather than a single "safe" number. Their findings confirmed 4% as a reasonable baseline for a 30-year horizon with a 50-75% equity allocation, showing success rates above 95%.
However, both studies shared a critical limitation: they relied exclusively on US market data from what turned out to be the strongest century for equities in global history.
| Country | Safe Withdrawal Rate (30yr, 95% Success) | Equity Real Return (1900-2020) |
|---|---|---|
| United States | 4.0% | 6.7% |
| United Kingdom | 3.4% | 5.4% |
| Japan | 2.3% | 4.1% |
| Germany | 2.6% | 3.2% |
| Italy | 2.1% | 2.5% |
| International Average | 3.0% | 4.5% |
Dimson, Marsh, and Staunton's international dataset reveals a sobering reality. The US experience was exceptional, not typical. For retirees in most developed markets, a 4% withdrawal rate would have failed historically. Japan and Italy produced safe withdrawal rates closer to 2%, reflecting extended periods of poor equity and bond returns.
Valuation Matters: The CAPE-Adjusted Approach
Wade Pfau's research demonstrated that the single best predictor of a sustainable withdrawal rate is the Shiller CAPE ratio at the time of retirement. When CAPE is low (below 12), historical safe withdrawal rates have exceeded 5%. When CAPE is elevated (above 25), the safe rate drops below 3%.
| CAPE Range at Retirement | Median Safe Withdrawal Rate | 10th Percentile (Worst Case) | Historical Frequency |
|---|---|---|---|
| Below 12 | 5.8% | 4.9% | 18% |
| 12-18 | 4.8% | 4.0% | 35% |
| 18-25 | 4.2% | 3.3% | 30% |
| Above 25 | 3.4% | 2.5% | 17% |
This finding has immediate practical significance. As of early 2026, the US Shiller CAPE ratio stands near 33, well above the historical average of approximately 17. Pfau's framework suggests a retiree entering the market at these valuations should plan for a withdrawal rate closer to 3.0-3.5% rather than 4%.
The mechanism is straightforward: high CAPE ratios predict lower subsequent 10-year equity returns. A retiree who withdraws 4% from a portfolio that earns only 2-3% real in the early years faces sequence-of-returns risk; the early drawdowns compound, making recovery increasingly difficult even if returns normalize later.
Dynamic Withdrawal Strategies: Adapting to Reality
The most significant advance in retirement income research has been the development of dynamic withdrawal strategies that adjust spending based on portfolio performance. Unlike the fixed withdrawal approach (where you increase the initial dollar amount by inflation regardless of what markets do), dynamic strategies explicitly trade withdrawal stability for portfolio longevity.
Guyton and Klinger (2006) proposed a decision-rule framework with three guardrails:
The Prosperity Rule: If the portfolio grows enough that the current withdrawal rate falls below the initial rate by more than 20%, the retiree takes a raise (increasing withdrawals by 10%).
The Capital Preservation Rule: If the portfolio declines enough that the current withdrawal rate exceeds the initial rate by more than 20%, the retiree takes a pay cut (reducing withdrawals by 10%).
The Portfolio Management Rule: In years when the portfolio declines, the withdrawal is not adjusted for inflation.
Blanchett, Kowara, and Chen (2012) at Morningstar tested an optimized dynamic framework and found that dynamic strategies could support initial withdrawal rates 15-30% higher than a fixed approach with equivalent portfolio survival probabilities.
| Strategy | Initial Withdrawal Rate (95% Success, 30yr) | Worst-Year Income Cut | Median Terminal Wealth |
|---|---|---|---|
| Fixed (Bengen) | 4.0% | 0% (inflation-adjusted) | $620K |
| Constant Percentage | 5.2% | -28% | $0 (by design) |
| Guyton-Klinger | 5.1% | -10% per rule trigger | $510K |
| Floor-Ceiling (80/120) | 4.8% | -20% from initial | $440K |
| CAPE-Adjusted | 4.6% | -15% from initial | $580K |
The tradeoff is clear. Dynamic strategies can support higher average spending, but they require willingness to accept income volatility. The Guyton-Klinger framework, for instance, triggered a pay cut in roughly 20-25% of historical years. For retirees with flexible spending (discretionary travel, dining) layered on top of fixed expenses (housing, healthcare), this structure maps naturally onto real spending patterns.
The Rising Equity Glide Path
Conventional wisdom holds that retirees should reduce equity exposure over time, shifting toward bonds as they age. Research by Pfau and Kitces suggests the opposite may be superior.
A rising equity glide path starts with a conservative allocation (perhaps 30% equity, 70% bonds) at retirement and gradually increases equity exposure to 60-70% over the first 15-20 years. The logic addresses sequence-of-returns risk directly: by holding less equity early in retirement (when the portfolio is largest and most vulnerable to drawdowns), the retiree reduces the probability of catastrophic early losses. As the portfolio survives and the time horizon shortens, the increasing equity allocation captures the equity premium over the remaining period.
| Glide Path Strategy | Initial Equity | Final Equity (Year 30) | Success Rate (4% WR) | Worst-Case Terminal Wealth |
|---|---|---|---|---|
| Declining (70→30) | 70% | 30% | 90.5% | -$42K (shortfall) |
| Static (50/50) | 50% | 50% | 92.1% | $12K |
| Rising (30→70) | 30% | 70% | 94.8% | $85K |
| Rising (20→80) | 20% | 80% | 93.2% | $62K |
The rising glide path improved success rates by approximately 4 percentage points relative to the conventional declining path in Pfau's simulations. More importantly, it substantially improved worst-case outcomes: the worst historical cohort under a rising glide path retained meaningful portfolio value, while the declining path produced a shortfall.
The Bucket Strategy: Behavioral Architecture
The bucket strategy segments the retirement portfolio into time-based tranches: a short-term bucket (1-2 years of expenses in cash/money market), a medium-term bucket (3-7 years in bonds), and a long-term bucket (8+ years in equities). While the bucket approach is mathematically equivalent to a total-return strategy with systematic rebalancing, it serves a critical behavioral function.
| Bucket | Time Horizon | Allocation | Purpose | Yield/Return Target |
|---|---|---|---|---|
| Short-term | Years 1-2 | Cash, Money Market | Immediate spending | 4-5% (current rates) |
| Medium-term | Years 3-7 | Investment-grade bonds | Income stability | 4-6% |
| Long-term | Years 8-30 | Diversified equities | Growth, inflation hedge | 7-10% nominal |
Research by Blanchett (2015) confirmed that bucket strategies produce equivalent mathematical outcomes to total-return approaches, but their behavioral architecture significantly reduces the probability of panic selling during bear markets. Retirees who can see two years of spending safely in cash are far less likely to liquidate equities at distressed prices. The behavioral alpha of this structure may exceed 1% annually in practice.
Historical Success Rates by Withdrawal Rate
The following table synthesizes the cumulative research, showing success rates across withdrawal rates, time horizons, and allocation strategies.
| Withdrawal Rate | 20-Year Success | 25-Year Success | 30-Year Success | 35-Year Success | 40-Year Success |
|---|---|---|---|---|---|
| 3.0% | 100% | 100% | 100% | 99% | 98% |
| 3.5% | 100% | 100% | 98% | 95% | 91% |
| 4.0% | 100% | 98% | 95% | 87% | 82% |
| 4.5% | 99% | 93% | 85% | 76% | 68% |
| 5.0% | 96% | 84% | 72% | 61% | 52% |
| 5.5% | 89% | 72% | 58% | 47% | 38% |
These success rates assume a 60/40 US portfolio with annual rebalancing. International data would reduce these rates by approximately 5-10 percentage points, and current high-CAPE conditions suggest additional caution.
Practical Synthesis
The research points to several conclusions that have survived multiple methodologies:
First, the 4% rule is a reasonable starting point for a US retiree with a 30-year horizon and moderate equity allocation, but it is not universally safe. International evidence, current valuations, and longer time horizons all argue for a lower initial rate of 3.0-3.5%.
Second, dynamic withdrawal strategies dominate fixed withdrawal approaches on nearly every metric. A retiree willing to accept 10-20% income variability can safely start with withdrawal rates 15-30% higher than the fixed approach.
Third, the CAPE ratio at retirement is the single most important variable in determining sustainable withdrawal rates. Retiring when CAPE is above 25 reduces the safe rate by approximately 1.5 percentage points relative to low-CAPE environments.
Fourth, rising equity glide paths counter-intuitively outperform declining paths by directly addressing sequence-of-returns risk, improving worst-case outcomes by meaningful margins.
Finally, the behavioral dimension matters as much as the mathematical framework. A theoretically optimal strategy that leads to panic selling during a bear market will underperform a simpler strategy that the retiree can maintain through adversity.
Related
Written by Elena Vasquez · Reviewed by Sam
This article is based on the cited primary literature and was reviewed by our editorial team for accuracy and attribution. Editorial Policy.
References
- Bengen, W. P. (1994). Determining Withdrawal Rates Using Historical Data. Journal of Financial Planning, 7(4), 171-180.
- Cooley, P. L., Hubbard, C. M., & Walz, D. T. (1998). Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable. AAII Journal, 20(2), 16-21.
- Guyton, J. T., & Klinger, W. J. (2006). Decision Rules and Maximum Initial Withdrawal Rates. Journal of Financial Planning, 19(3), 48-58.
- Pfau, W. D. (2012). Capital Market Expectations, Asset Allocation, and Safe Withdrawal Rates. Journal of Financial Planning, 25(1), 36-43.
- Blanchett, D., Kowara, M., & Chen, P. (2012). Optimal Withdrawal Strategy for Retirement Income Portfolios. Morningstar Investment Management Working Paper.
- Dimson, E., Marsh, P., & Staunton, M. (2021). Credit Suisse Global Investment Returns Yearbook 2021. Credit Suisse Research Institute.
- Kitces, M. E., & Pfau, W. D. (2015). Retirement Risk, Rising Equity Glide Paths, and Valuation-Based Asset Allocation. Journal of Financial Planning, 28(3), 38-48.
- Blanchett, D. (2015). The ABCDs of Retirement Success. Journal of Financial Planning, 28(5), 30-39.