The Military Guide


Is the 4% withdrawal rate really safe?

by Doug Nordman


What if 4% is all wrong?

By “4%” I mean the method tested by the Trinity Study over a decade ago.  It determined a portfolio would survive for 30 years if withdrawals started at 4% of the portfolio’s initial value and then were raised each year by the rate of inflation.

Skeptics point out that the Trinity Study is based on a narrow slice of U.S. stock market history. It’s possible that 20th-century returns were unusually high and would never be repeated, so an entire generation of retirees may be about to risk their assets (and retirements) on a stock-market fluke.

The 4% withdrawal system isn’t a realistic model of retiree spending, either. Retirees do not blissfully spend 4% and raise it every year no matter what.  People tend to spend less when the economy is bad and more when the economy is good. Older retirees may spend less after age 80 while younger retirees will want to spend more in their 40s and 50s.

Finally, critics have pointed out that almost all retirees have a safety net: Social Security. This also changes their spending behavior because they know that some of their portfolio will be augmented by an inflation-fighting annuity. Even retirees who aren’t eligible for Social Security should annuitize a portion of their retirement portfolio to make sure that they have a diversified source of income.


In December 2010, Professor Wade Pfau studied the stock-market returns of other countries using the same methods as the Trinity researchers.  Pfau’s results were sobering: the 4% rule failed in 17 other developed countries.  In other words, the 4% rule was an American anomaly.

Just 10 months later, Professor Pfau is back with a different retirement approach.  He looks at two problems with mutually opposing solutions:

The economic term for these problems is “maximizing utility”. Retirees want to maximize their spending while understanding how long they might have to live without their portfolio if the higher spending rate causes it to run out.

One issue with “zero failure” is that it drives the “safe” withdrawal rate down to an extremely conservative number– and it’s unrealistically conservative. Historical retirement calculators don’t have enough stock-market performance data to generate a statistically valid prediction of a portfolio’s failure rate. However they do have performance data from two World Wars, the Great Depression, and a long period of underperformance (1966-1982).  William Bernstein’s “Retirement Calculator from Hell” articles claim that this is the equivalent of a retirement success rate of 80%, and that retirees can’t realistically demand a higher success rate. Of course that implies they’d also have to be willing to risk a 20% failure rate. Pfau’s international research suggests they might have an even higher risk if they’re living outside the U.S.

So how do retirees maximize their retirement portfolio’s utility while handling the risk of higher failure rates? The first step is annuitizing a portion of the portfolio for a subsistence lifestyle. For many retirees, this is Social Security. For others it may be a higher number– for example an insured income of $24,000/year from a single premium immediate annuity, perhaps including an inflation adjustment.

This changes the retirement risks. Retirees will never completely run out of money (as long as Social Security or the annuity are paying out). Even if their retirement portfolio fails near the end of their lives, they can drastically cut back their lifestyle to live within their annuity income. Instead of “failure”, retirees can focus on how long they might have to live on a bare-bones budget after their portfolio runs out.

Now the problem can be stated more practically: What’s the lowest income a retiree would be willing to live on if the rest of their portfolio runs out? How long would they be willing to live that way? What withdrawal rates and asset allocation could maximize their spending while minimizing the risk that they’d have to live on just their annuity income?

This gives retirees the choice of investing more aggressively or perhaps annuitizing more of their income. In either situation they’d be able to spend a higher percentage of their remaining portfolio– at least 4% and as much as 7%– knowing that the “worst case” would be having to live on just their annuity income. Of course many retirees might still prefer to start with 4% of a $1M portfolio (plus a small annuity) instead of 7% of a $500K portfolio (backed by a larger annuity). However they now have a realistic choice, and they can put numbers on it.

How did retirees choose among these situations? Most of them opted to raise their portfolio spending during the earlier part of their retirement, even though they were spending at least the “4% rule” while raising the risk that their portfolio would run out. They were willing to live with that risk because their own “worst case” would be living on their annuitized income. Even if they lived longer than their statistical life expectancy, they still had income– their “mortality risk” was covered by the annuity.

If you’re a military retiree, then you have one of the world’s best annuities. Even if you’re a military veteran with no pension, you’re still eligible for some Social Security benefits.

Pfau’s study goes a long way toward a realistic model of retirement spending, but it still neglects one aspect: it assumes that retirees keep spending until their retirement portfolio is all gone. The reality is that retirees will still cut back their spending when their portfolio is depleted by a bear market. In other words, instead of waiting until their portfolio runs out to start living on their annuity income, they might do it earlier in their retirement for a year or two. This means that they can still decide to cut back their spending to help their portfolio survive. The difference is that now they know how much they’d have to cut back their spending, and can maybe even predict how long they’d have to live that way.

Getting back to the original question: is the 4% withdrawal rate really safe? Yes: if you annuitize a portion of your income with Social Security, a military pension, or even a single-premium immediate annuity.

Now retirees have a practical retirement plan, one that minimizes the effect of failure while maximizing their retirement spending. Better yet, the solution improves their lifestyle without handing over large sums of money to their beneficiaries or their insurance company.

This retirement still can’t be completely modeled in a simple retirement calculator, but retirees can use different calculators to come up with a very good answer. In a later post we’ll talk about retirement calculators that help implement the plan.

I’ll also talk about USAA’s latest retirement-planning website tools. Yesterday I interviewed Zack Gipson, USAA’s executive in charge of goals-based planning calculators. If you’re a veteran who’s served honorably (or a family member) then you’re eligible for USAA membership.  Even if you’re not eligible for membership then you can still purchase USAA’s investment products– and it’s worth it just to use these tools.

Related articles:
Details of the 4% Safe Withdrawal Rate
Problems with retirement calculators
Trends in retirement
Military retirement: how much can I really spend?
Military retirement spending: how much will I need?
Retiring without a military pension
How many years does it take to become financially independent?

Does this post help? Sign up for more free military retirement tips via e-mail, Facebook, or Twitter!