Skip to Content
Rekenthaler Report

Is 3% the New 4%?

The news on retiree withdrawal rates is better than it seems.

Half Full
There's been much discussion about the rule of thumb that retirees can safely spend 4% of their initial assets each year during retirement, adjusted for inflation. That is, somebody who retires with a $500,000 nest egg can withdraw $20,000 in year one, $20,000 adjusted for year one's inflation rate in year two, and so forth. The 4% rule has been in force since the early 1990s, when a rocket engineer (and financial planner) named Bill Bengen published an article that popularized the notion. Recently, the rule has come under attack.

Earlier this year, both The New York Times and The Wall Street Journal pronounced the 4% rule to be on its deathbed. The problem is the combination of low bond yields and high stock valuations. With 10-year U.S. government bonds yielding less than 3%, there's no way to achieve a 4% payout without assuming either a large amount of credit risk or expecting equities to carry the weight. The former is not acceptable, and the latter is not credible. Consequently, some now argue that 3% is the new 4%.

Well, yes...but no. As Morningstar's David Blanchett points out, Bengen created the 4% rule as a floor, not as a ceiling. The original study sought a withdrawal rate that always worked over a 30-year time horizon. "Always" is a high percentage, and 30 years is a long time. The study also used only two asset classes: large-company U.S. stocks and intermediate-term U.S. government bonds. (When Bengen added a third asset class, small-company U.S. stocks, he raised the acceptable withdrawal rate to 4.5%.)

Thus, even if the acceptable withdrawal rate per the study's approach has subsided to something close to 3%, it doesn't follow that a retiree must settle for such a low figure. Three simple changes can significantly improve the rate.

1. Be Flexible
Studies typically use rigid models, because building a flexible model is time-consuming, and interpreting the results is trickier, too. (It's a lot easier to explain "4% per year" than it is "4% per year, but adjust for circumstances, so that sometimes the rate might be less than 4% and other times the rate might be higher than 4%, depending on market results and the time remaining in the simulation.") Retirees needn't be so structured. There's no need to lock in to a fixed multidecade plan that does not deviate in response to market movements.

The first few years during the retirement withdrawal period are critical. An early bear market raises the effective withdrawal rate as the asset pool shrinks. (Recall that the withdrawal amounts are established by the size of the initial balance, not the ongoing balance.) Imprudent withdrawals can push the portfolio past the tipping point, wherein the asset pool is too small to withstand the long-term spending demands.

In such cases, it's prudent to tighten one's belt, take out less money, and wait for a market recovery. This approach is akin to cutting back spending during the accumulation period when income declines.

2. Choose a Lower Success Percentage
Why plan for 100% success? Guaranteed success in a model doesn't mean guaranteed success in reality. The real world brings uncertainties that are outside the scope of any model. The seeming comfort of having a plan that never fails is an illusion. No such plan exists.

Aiming for 100% success via a model of fixed spending is an arbitrary--and overly conservative--decision. Might as well aim lower for something that permits a more generous withdrawal rate. My initial thought is 90%, but David Blanchett urges something closer to 80%. After all, accepting a 10%-20% risk per the model at the outset doesn't mean accepting that amount of risk throughout the withdrawal period. As discussed, if the markets perform poorly, the retiree can cut back, thereby turning an unlucky draw from the model's perspective into a successful real-world outcome.

3. Choose a Shorter Time Horizon
Only one 65-year-old man out of eight will live to the age of 95. The figures rise for women, and rise yet again for at least one spouse in a marriage, but it is always distinctly a minority possibility. What's more, most 95-year-olds spend relatively little money, aside from health care, which is something of a different issue. They aren't strong and healthy enough to do many discretionary activities.

For many, a 30-year time horizon is too long. Consider the case of the 65-year old single male, who has a 12% chance of surviving to age 95. Of those who do make it to age 95, arbitrarily we'll say that one in three have the health to engage in discretionary spending. That's one man in 25 who truly needs the full 30-year horizon.

Also, it's not as if planning for a 25-year horizon means running out of money the day that horizon ends. On the contrary, in a model with a 25-year horizon and a 80% success rate, most plans survive to 30 years. Once again, it's worth remembering that the 4% rule was built to handle the very worst case--not the good cases, nor the average situation, nor even the relatively bad instances. Rather, the very single worst case.

Of course, minimizing costs is also critical, even more than during the accumulation phase. Various investment strategies are also helpful, including a diverse asset allocation that includes alternative investments (if they are low-cost!); tactical tilts away from popular, expensive assets; and possibly purchasing fixed annuities. Those topics, however, are outside the scope of this column.

What They Don't Teach You in Business School
From Sports Illustrated, via Morningstar's Russ Kinnel: "When you have an investment base because you're publicly traded, the good thing is, if someone doesn't like what you're doing, they sell their stock. You're not beholden to them in any way," said Grand Canyon CEO and president Brian Mueller. That is a great quote--I need to stop buying that magazine only for the pictures.

John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.