Bill Bengen Boosts the “4% rule” to 4.7%

That’s why Nobel laureate economist William Sharpe says the “nastiest, hardest problem in finance” is distributing your assets to yourself after a lifetime of saving.

Thankfully, Bill Bengen has made things a little easier in his new book, “A Richer Retirement: Supercharging the 4% Rule to Spend More and Enjoy More” (Wiley, August 2025, 272 pages).

Bengen is an MIT-educated rocket scientist, retired financial advisor, and somewhat unintentional founder of “the 4% rule.” The widely used rule of thumb resulted from a study he did more than 30 years ago.

The new book is an update of this study. Bengen has added asset classes to his original work, and now advocates considering market valuation (the Shiller PE) and expected inflation alongside other withdrawal factors.

Ultimately, greater portfolio diversification has allowed him to increase the amount one has been able to withdraw safely from 4.1% in the past to 4.7%.

Early Effort

As a financial advisor in the early 1990s, Bengen found himself flummoxed when his retired clients would ask him how much they could spend. His first study showed a portfolio of 60% large-cap U.S. stocks and 40% intermediate-term U.S. government bonds, rebalanced annually, could withstand an initial distribution of 4.15% in the worst instance. That assumed an inflation adjustment on the initial dollar value of the first distribution in subsequent years for a full three decades.

Bengen made calculations for a hypothetical person retiring every quarter from the start of 1926. He assumed a tax-advantaged account such as a 401(k) or IRA so as not to have to consider a portfolio being taxed at different rates for capital gains, dividends, and interest income. Of course, distributions from such accounts are taxed, but not having to make tax adjustments to the account itself kept the study cleaner. In any case, his study was understood as advocating the “4% rule.”

He didn’t quite intend to impose that distribution rule on anyone, but acknowledged it as the number that had never failed in his study — the maximum safe or “SAFEMAX” number. Failure meant exhausting all of one’s savings before 30 years.

That worst case (the one that was successful only with a 4.15% or less initial withdrawal rate) was based on retiring in October 1968 because the first years of that retirement scenario imposed both bad market returns and inflation on an investor.

The Great Depression also imposed bad market returns on investors, but did so with the benefit of deflation, meaning hypothetical retirees adjusted their payments to themselves downward rather than upward in subsequent years.

Diverse Developments

There is significant variability in Bengen’s results.

Many of the hypothetical quarterly retirees in his studies could withdraw substantially more than 4.15%, and “the average SAFEMAX for all 349 retirees [in the new study was] approximately 7.1%,” according to Bengen. But success depended on good market returns, especially in early years of retirement, and controlled inflation.

A 4.7% initial withdrawal rate may be the new SAFEMAX, but variability remains. At a 6% withdrawal rate, many portfolios lasted 50 years or more, while some didn’t make it to 20 years. Also, a 6.75% withdrawal rate was successful a little more than half of the time.

The variability of a safe withdrawal rate in Bengen’s study crystalizes Sharpe’s comment about the difficulty of distributing assets.

It can mean a lot of people wind up being too conservative in how much they withdraw. “For many….retirees, the universally ‘safe’ rate was too safe,” Bengen emphasizes, adding that “they cheated themselves by defaulting to the Universal SAFEMAX.”

But it also means it’s hard to blame them given how catastrophic withdrawing too liberally is. And there may also be moments in the future during which 4.7% fails. There are no guarantees that the past century of market history and the past SAFEMAX will hold for the future.

Adding Asset Classes

In his new study, Bengen uses a 55%/45%/5% allocation — with the 5% representing U.S. Treasury bills — and incorporates U.S. small-cap stocks, U.S. micro-cap stocks, U.S. midcap stocks, and international stocks into the equity portion.

Adding U.S. small-caps alone bumped the original 4.15% SAFEMAX rate to 4.5%, while all the others added only 0.20 percentage points to SAFEMAX.

Bengen hints that additional assets classes could help a bit more, but also adds, “I suspect we are now in the realm of diminishing returns.”

Up until recently, I believed TIPS could help boost the SAFEMAX given the fact that many of them have been trading to deliver inflation (or CPI) plus 2% to investors who hold them to maturity.

This was not the case for most of 2020 and 2021, when they were priced to deliver less than CPI to their owners.

However, the recent replacement of the head of the Bureau of Labor Statistics means there could be political pressure to suppress CPI. That, in turn, could cause TIPS, which experience accretion tied to CPI, to be tethered to a phony inflation gauge.

Price Points

Besides adding more asset classes to model portfolios, Bengen’s other innovation is a two-factor model consisting of a stock market evaluation and an inflation estimate. The stock market assessment is the cyclically adjusted PE (CAPE) Ratio or Shiller PE. The metric evaluates the market by comparing its price level with the past decade’s worth of the market’s real average earnings.

Using his existing data on retirees, Bengen finds that a high CAPE combined with a high inflation regime lowers the SAFEMAX.

He even provides helpful tables suggesting a SAFEMAX depending on starting CAPE and estimated inflation.

Readers should note that the S&P 500’s CAPE has risen to much higher levels than it was for Bengen’s last group of retirees with full 30-year periods. However, the CAPE is not quite as high as it was in early 2000, and that period — though it hasn’t lasted a full 30 years yet — did not “break” the new 4.7% rule.

What typically breaks a retirement distribution plan, Bengen says, is what happens in the first 10-12 years, and the 2000 and 2007 retirees have passed that crucial period successfully.

And regarding today’s retirees, Bengen doesn’t advocate the “Universal SAFEMAX.” He says, “Typically, a SAFEMAX between 5.25% and 5.5% appears approximately correct. I see no need to go as low as the ‘Universal SAFEMAX of 4.7% unless inflation again becomes a serious problem.”

In other words, there are two prices that affect SAFEMAX — security prices and consumer prices. When each gets high, it threatens to lower SAFEMAX. But Bengen thinks consumer price inflation rather than asset price inflation has the better chance of breaking a retirement plan.

As of this review, the CAPE is at 38, not far from its highest-ever 44 reading in early 2000. Profit margins are also high, and seem to justify a higher CAPE, which has averaged a breathtaking 27 since 1990, much higher than its longer-term 17 since 1880.

Retirees must consider whether profit margins, stock prices, and valuation multiples have reached a permanently high plateau, and perhaps even if that represents a problem for capitalism itself.

Additionally, inflation — almost anyone would say — appears to be more of a wild card now than it was in 2000, with a generation or two of deflationary labor offshoring being called into question. Of course, the advance of artificial intelligence could provide still more deflationary pressure, nullifying any inflation resulting from deglobalization.

In the end, Bengen has not refuted Sharpe’s assertion: distributing your assets in retirement is a difficult business. But Bengen’s book provides retirees and their advisors with all they could hope for: the right questions and — if not precise answers — useful suggestions, tables, and guidelines.

A message from Advisor Perspectives and VettaFi: To learn more about this and other topics, check out some of our webcasts.

More Active Fixed Income Topics >