CAPE and the 4% Rule: What 145 Years of Market History Say About Retiring Now
August 1, 2026
The historical record links expensive markets with lower safe withdrawal rates. At today's CAPE, however, the completed evidence runs out before the valuation does.
American stocks are in one of the stronger bull markets on record. The S&P 500 closed above 7,000 for the first time in April, setting a new all-time high.[1] Whether that is good news depends on what you are paying for it, and one widely used yardstick for that question is Robert Shiller's cyclically adjusted price-to-earnings ratio.[2]
What does CAPE represent? A high CAPE means investors are willing to pay more for a dollar of smoothed real earnings, either because they expect earnings to grow well past that ten-year average or because they are content to accept a thinner return for carrying the risk. Sometimes that confidence is vindicated: earnings grow into the price. Sometimes the growth that was priced in never arrives, and the price falls back to meet the earnings instead. The ratio cannot tell you which one we are in. It is better used as a long-horizon risk indicator than as a market timer; Campbell and Shiller's original work connected smoothed earnings measures with the predictability of long-horizon returns.[3]
Let's look at the historical CAPE by year:
CAPE across market history
Annual January observations show how unusual today's starting valuation is, and how long elevated readings can persist.
As shown in the chart, an elevated CAPE does not mean a collapse is around the corner. The 1987 crash followed a below-median January CAPE. But several difficult market eras began with high CAPE: 1929 at 27, 1966 at 24, and 2000 at 44, against a long-run median of 16.6.[2] When Alan Greenspan asked whether "irrational exuberance" had unduly raised asset values in December 1996, CAPE was about 28 and the market almost doubled before finally falling.[4][2]
The safe withdrawal rate
A safe withdrawal rate (SWR) is the percentage of an investment portfolio withdrawn in the first year of retirement, with the withdrawal amount then adjusted for inflation, while keeping the risk of depletion acceptably low over a chosen horizon. William Bengen's 1994 study found that an initial withdrawal of about 4%, followed by inflation-adjusted withdrawals, had survived every 30-year period in his historical US dataset.[5]
With that in mind, it is worth asking whether CAPE has any relationship with the safe withdrawal rate in retirement planning.
Relationship between CAPE and safe withdrawal rate
The interactive chart below explores the relationship between the market’s valuation at the start of retirement, measured by the cyclically adjusted price-to-earnings ratio (CAPE), and the withdrawal rate that a retirement portfolio could historically sustain. Use the controls to choose the CAPE at retirement, your initial withdrawal rate, and the length of retirement you want to test. The scatter plot then compares historical periods with similar starting CAPE levels. Each dot represents a retirement starting year: filled dots show plans that lasted for the selected period, while outlined dots show plans that ran out of money early. Select any dot to inspect that historical path in detail, including how the portfolio balance changed over time and the highest withdrawal rate that would have lasted.
Unless you change the controls, this back-test uses a 60/40 US stock/bond portfolio, annual rebalancing, January start dates, and start-of-year withdrawals that rise with inflation. Bond returns use a long-term Treasury proxy derived from Shiller's interest-rate series; fees and taxes are set to zero.[2] The 30-year cohorts overlap, so they are historical cases rather than independent trials. The results are educational and are not personal investment advice.
CAPE & withdrawal rates
No completed cases match CAPE 38–44 for this horizon.
Retiring in 1881, the hardest start in this range
For this 1881 start at CAPE 18.5, the highest rate that lasted was 6.7%. Your 4.0% plan lasted all 30 years.
Assumptions and method
Inflation-adjusted spending, annual rebalancing, start-of-year withdrawals, and no taxes.
Complete cohorts only: 30-year starts run through 1996. Overlapping starts are not independent trials.
Shiller U.S. market data through 2025. Educational only.
What the scatter plot tells us
1. Overall relationship
The 'cloud' gently slopes downwards, showing a negative relationship between CAPE and safe withdrawal rate (i.e. higher CAPE at retirement is generally associated with a lower historically sustainable withdrawal rate.)
2. Large variance at the same CAPE
Points with similar CAPE values still have noticeably different outcomes. CAPE matters, but it does not determine the result by itself.
3. Worst-case outcome within the selected range
The lowest point in the highlighted CAPE band identifies the most difficult historical starting period and therefore the withdrawal rate needed to survive every completed case in that band.
4. Limits of CAPE
The overlap and scatter make clear that CAPE is better viewed as a risk indicator than as a precise withdrawal-rate calculator.
Scaling the rate to the price
If the floor falls as CAPE rises, one possible policy is to set the starting rate as a function of the starting valuation. Karsten Jeske has extensively explored this form of CAPE-based withdrawal rule in his Safe Withdrawal Rate series.[6]
- w0
- initial withdrawal rate at retirement, as a share of the starting portfolio
- a
- base withdrawal rate, independent of the starting valuation
- b
- weight applied to the earnings-yield component
- CAPE
- cyclically adjusted price-to-earnings ratio at retirement
where 1/CAPE is the cyclically adjusted earnings yield, a rough proxy for expected real stock returns. There is no universally agreed pair of coefficients derived from financial theory. Jeske illustrates the rule with an intercept of 1.75% and a slope of 0.5, while using the more cautious 1.5%/0.5 combination in some examples.[6][7] His published rule is dynamic, recalculating withdrawals as the portfolio and CAPE change. In this article, we use the same formula only to set the initial withdrawal rate; subsequent withdrawals follow the backtest's inflation-adjusted spending path.
CAPE-scaled withdrawal policy overlay
Using the same method as above, we now overlay the CAPE-scaled withdrawal policy on top of the historical cohort data. The solid green line shows the CAPE-scaled withdrawal rate.
From the chart, we can see that the formula is a valuation-responsive policy. CAPE-scaled policy allows substantially higher withdrawals at low valuations. That may prevent excessive underspending when markets are cheap, and vice versa. However, the formula does not guarantee the historical floor (and thus the reader can fine tune the parameter to their liking). Hence, it does improve the 'efficiency' but does not guarantee the safety of the withdrawal rate.
Conclusion
The historical record suggests that starting valuation matters: higher CAPE levels have generally been associated with lower sustainable withdrawal rates. But CAPE is not a precise forecasting tool, as we've seen from the chart above, and similar starting valuations have produced very different retirement outcomes.
A valuation-scaled rule such as 1.75%+0.5/CAPE can make a withdrawal policy more responsive to market conditions, allowing more spending when valuations are low and less when they are high. It does not, however, guarantee safety—especially at today’s CAPE, which lies beyond the range of completed 30-year historical cohorts.
The practical takeaway is simple: the 4% rule remains a useful benchmark, but at unusually high valuations, retirees may want a lower starting rate, more spending flexibility, or both.
Sources
- Associated PressHow major US stock indexes fared Wednesday 4/15/2026April 15, 2026. https://apnews.com/article/wall-street-stock-dow-nasdaq-b8818e64043e32c2e5bae57213d1a7ee
- Shiller DataUS Stock Markets 1871–Present and CAPE RatioRetrieved August 1, 2026. https://shillerdata.com/
- National Bureau of Economic ResearchStock Prices, Earnings and Expected DividendsFebruary 1988. https://www.nber.org/papers/w2511
- Board of Governors of the Federal Reserve SystemCentral Banking in a Democratic SocietyDecember 5, 1996. https://www.federalreserve.gov/boarddocs/speeches/1996/19961205.htm
- Journal of Financial PlanningDetermining Withdrawal Rates Using Historical DataOctober 1994. William P. Bengen. https://www.financialplanningassociation.org/sites/default/files/2021-04/MAR04%20Determining%20Withdrawal%20Rates%20Using%20Historical%20Data.pdf
- Early Retirement NowThe Ultimate Guide to Safe Withdrawal Rates – Part 18: Flexibility and the Mechanics of CAPE-Based RulesAugust 30, 2017. https://earlyretirementnow.com/2017/08/30/the-ultimate-guide-to-safe-withdrawal-rates-part-18-flexibility-cape-based-rules/
- Early Retirement NowDealing with a Bear Market in Retirement – SWR Series Part 37March 25, 2020. https://earlyretirementnow.com/2020/03/25/dealing-with-a-bear-market-in-retirement-swr-series-part-37/
About the author
Roen is a Fellow of the Society of Actuaries (FSA) and a Chartered Enterprise Risk Actuary (CERA) working in life insurance. His work focuses on Solvency II, capital management, and asset–liability management, with deep experience in financial and stochastic modelling. On this site, he uses the same actuarial tools applied in insurers to help individuals think more rigorously about retirement and long-term financial risk.