A A I I J o u r n A l M A R C H 2 0 2 2 2 1
A A I I . C O M / J O U R N A L
P O R T F O L I O S T R AT E G I E S
The 4% rule is based on U.S. historical data, and we’ve
never had interest rates this low. Four percent is supposed
to be a safe withdrawal rate. I don’t think in this kind of
interest rate environment it can be considered a safe with-
drawal rate if you’re otherwise trying to maintain all of the
rules basic assumptions: a 30-year retirement, no invest-
ment fees, no taxes and a fairly aggressive asset alloca-
tion of 50% to 75% stocks. It’s just going to be under more
strain than the U.S. historical data would suggest.
[See Figure 1 for portfolio success rates using the 4%
rule.]
Is there a safe number now that you would give out?
There was a Morningstar study that was recently
updated. I was a co-author on the previous iteration, and
I think we had a number in the ballpark of 2.8%. The
updated study came out with a 3.3% withdrawal rate. Any-
thing in the area around 3% is more realistic in this inter-
est rate environment.
A New Perspective
on Withdrawal and
Allocation Strategies
for Retirees
Factoring in worst-case scenarios for your
early years of retirement helps ensure your
portfolio lasts.
AN INTERVIEW WITH WADE D. PFAU, PH.D.,
CFA, RICP
Wade Pfau is a widely followed retirement researcher and
a professor and program director at The American College of
Financial Services in King of Prussia, Pennsylvania. He is also
principal and director for McLean Asset Management and
RISA LLC. In this first of a two-part interview, we discuss both
retirement portfolio withdrawal rates and allocation strategies
for retirees.
—Charles Rotblut, CFA
There is much debate about whether the traditional rule
of taking 4% of a retirement portfolio’s starting value and
adjusting it annually for inflation still makes sense. What
are your thoughts?
Theres a lot of assumptions that go into calculations
for the 4% rule, and some of them are not very realistic.
One in particular is the notion that you never adjust your
spending in response to market performance, which actu-
ally creates more sequence risk. But if you’re trying to
stay as close as possible to the core assumptions, I think
a 4% withdrawal rate is too high at the present, primarily
because of the low-interest-rate environment.
Wade D. Pfau, Ph.D., CFA, RICP, is the
program director of the Retirement Income
Certied Professional designation and
a professor of retirement income at The
American College of Financial Services in
King of Prussia, Pennsylvania. He is author
of the book “Retirement Planning Guidebook” (Retirement
Researcher Media, 2021). Find out more at www.aaii.com/
authors/wade-pfau.
FIGURE 1
Portfolio Success Rates for the 4% Rule
This chart shows the likelihood of a retiree not outliving
their portfolio over a 30-year period when following the
William Bengen 4% withdrawal rule and using a portfolio
allocation of 50% stocks and 50% bonds. The historical
simulations use average returns for rolling periods. The
Monte Carlo simulations randomize the sequence of
returns. The adjusted returns scenario assumes a real
(inflation-adjusted) return of –1% for intermediate-term
bonds.
Source: “Retirement Planning Guidebook,” by Wade Pfau (Retirement
Researcher Media, 2021).
© 2022 by the American Association of Individual Investors, 625 N. Michigan Ave., Chicago, IL 60611; (312) 676-4300.
A A I I . C O M / J O U R N A L
2 2 A A I I J o u r n A l M A R C H 2 0 2 2
It was the best-case scenario in history, but it was too
late for those earlier retirees to benefit from what hap-
pened in the second half of their retirement. It was really
what happened in those early retirement years that led to
these worst-case scenario outcomes. The early retirement
years matter much more than the later retirement years in
determining retirement sustainability.
[See Figure 2 for an illustration of sequence of returns
risk.]
Is there a benchmark retirees can use to determine when
their portfolio is on safe ground if they reach it?
The thing about the 4% rule method is that the initial
withdrawal rate only applies in the first year of retirement.
Afterward, you’re always just adjusting your spending with
inflation, but you’re not worried about what the current
withdrawal rate is.
You could simply divide current spending by your
remaining portfolio balance to get a new withdrawal rate.
If things are going well in retirement and you’re spending
conservatively, your actual current withdrawal rate could
decline throughout retirement.
If your portfolio balance ever falls below where it
started just in nominal terms, without even making infla-
tion adjustments, that’s kind of a signal that you may be
getting into trouble. So, you might want to start cutting
back on spending and prepare for the possibility that you
are in one of those scenarios where you may deplete your
investment portfolio, or at least have an uncomfortable
time later in retirement.
Some investors prefer using time-oriented buckets. Could
you share some of the weaknesses and advantages of doing
a bucket strategy versus a traditional allocation?
With bucketing, the advantages are going to be behav-
ioral. And in some recent research I’ve done with Alex
Murguia, we talk about retirement income styles and iden-
tify bucketing or time segmentation as one of four viable
retirement income styles for people who do want some sort
of contractual protection, but also want a lot of optionality.
They use bonds to cover upcoming expenses, and then that
creates a window for not having to sell their stock market
investments during a market downturn.
Retirees can just continue to spend down their fixed-
income buckets. If that helps them to stay the course with
their strategy, then that’s a sort of behavioral advantage
toward having a better retirement outcome.
Now, in terms of disadvantages, this is a strategy where
it’s often discussed at a basic level but without details
about when you should refill your buckets. So, if you have,
just for example, five years’ worth of projected expenses in
bonds and everything else in stocks, you need some sort of
rule. As I spend my bonds, when am I going to sell stocks to
In my analysis of worst-case scenarios, it seemed that
a retiree was most likely to outlive their savings when a
sequence of bad returns occurred early in retirement. Have
you noticed something similar or different?
The worst-case scenario is some sort of prolonged
downturn in the early retirement years. If inflation is high,
that does seem to require lower withdrawal rates as well.
The worst case in the U.S. historical data was retiring
in the 1960s when there were big downturns in the stock
market, like in 1966. In 1973 and 1974 the S&P 500 index
fell by 47%. Combined with the inflation rate at that time,
it set up a scenario where these early market downturns
disrupted retirement.
In the data that William Bengen (the creator of the 4%
rule) used, 1966 was the retirement year that led to the
worst-case scenario. The year 1982 was a big turning point.
After 1982, the markets did great and accounted for almost
half of a 30-year retirement for the 1966 retiree. If youd
retired in 1982, you could have used almost a 10% with-
drawal rate.
FIGURE 2
Lifetime Sequence of Return Risk
Returns in the early part of an investor’s career have very
little impact on the absolute level of wealth accumulated
at the end of the savings period. As retirement nears and
a large amount of wealth has been accumulated, a given
percentage return (be it positive or negative) has an in-
creasing impact. Investors are most at risk around the date
of retirement as they have a large amount of wealth saved,
leading to proportionately large changes in wealth for a
given return.
Source: “Retirement Planning Guidebook,” by Wade Pfau (Retirement
Researcher Media, 2021).
A A I I J o u r n A l M A R C H 2 0 2 2 2 3
A A I I . C O M / J O U R N A L
If you just hold the individual bonds to maturity, you’ve
got the right duration for your expenses, because as the
bond matures, it covers your spending. You could have
paper losses if interest rates go up, but if you hold your
bonds to maturity, you know what you’re getting.
If you’re trying to replicate this with bond funds, yes,
it’s mathematically possible but it’s just a lot more compli-
cated to do in practice.
You also found that the total-return strategies had a
better outcome than bond ladders, particularly if you’re
increasing your allocation to equities once you’re retired. Is
that correct?
When I looked at comparing the time-segmentation
strategies to total-return strategies, I usually just look
at a simple kind of 60% stock/40% bond allocation for
the total returns, where you have a fixed asset allocation
throughout retirement. But using time segmentation
actually leads to a rising equity glide path in many cases.
So, a kind of U-shaped lifetime stock allocation.
The reason time segmentation can work is because
it can give you a rising equity glide path during troubled
market environments. You’ve got to be comfortable with
that. If you’re not, then it’s not necessarily going to work
for you.
If you automatically refill your bond bucket every year,
you can have a declining equity glide path in retirement
because as the portfolio gets into trouble, your stock piece
is getting smaller, but you keep selling from it to replenish
your bonds until at some point you don’t have stocks any-
more. This could leave you in a scenario where you may
run out of money.
You get to the end where your stocks are gone, and now
all you have left is a five-year bond bucket. If you spend
that down over the following five years, you’ll be out of
money. That’s where it doesn’t work as well.
Where bucketing does work is if you use a rising equity
glide path. When Michael Kitces and I first wrote about the
rising equity glide path (“Reduce Stock Exposure in Retire-
ment, or Gradually Increase It?,April 2014 AAII Journal),
we were thinking in terms of a total-return strategy, but it
shows up as well in a version of time segmentation that
works. It’s a rising equity glide path.
That leads to my next question about glide paths. Could
you explain to our members what a glide path is and why a
U-shaped path makes sense?
With a target-date fund, you have a higher stock allo-
cation when you are young. As you get to retirement, you
have a lower stock allocation. Target-date fund allocations
vary quite a bit, but they usually have between 20% and
40% in stocks at the retirement target date. That’s coun-
tered by all of the financial planning style research that
replenish my bonds so that I still have five years of bonds to
cover my upcoming expenses?
If you have a rule in place to automatically refill the
bucket every year, that can increase sequence of returns
risk, because you’re having to effectively take a full year
of spending out of a smaller stock portfolio to replenish
the bond portfolio. It can cause you to run out of money
sooner than if you just use a basic total-return strategy for
retirement. The only way bucketing can actually help get
a better outcome is the logic of not selling stocks during a
downturn.
The way bucketing is sold as an idea is to let my bond
ladder get smaller and smaller until at some point it could
even be gone. But if a stock market downturn occurs once
my bond ladder is depleted, I’ve already shifted toward
100% stocks.
A bucket strategy requires people to have a dynamic
asset allocation. It requires investors to effectively increase
their stock allocation during the market downturn. It’s not
that they’re buying more
stocks, it’s just that they’re
spending bonds and not
touching their stocks. This
pushes them more toward a
higher stock allocation. So,
if they’re comfortable doing that, then it’s good. If they’re
not comfortable doing that, if they use some other strategy
for refilling their short-term buckets, it could actually be
worse in absolute terms than just using a basic total-return
investing strategy.
Interesting. You wrote about bond ladders in your book,
“Retirement Planning Guidebook.Can investors use bond
funds if they don’t want to ladder individual bonds? How
important is it to have bonds maturing on certain dates as
opposed to just having a source of cash flow available in the
shorter-term bucket?
There are debates about this. I think it’s much easier to
explain bucket strategies with a bond ladder tied to when
you’re planning to spend the money. You could do it as
bond funds, but I think it’s harder than is commonly appre-
ciated. The idea is if you have a constant duration—bond
funds with a duration that matches your expenses—then
it shouldn’t matter.
But when you’re spending from the portfolio, the math
to have the right duration on your remaining bonds is
incredibly complicated. There are a few commercial pro-
viders who’ve created solutions for that, where they use
bond funds in lieu of having a bond ladder. But for a house-
hold trying to do that on their own, good luck. You really
need to be an advanced engineer to handle getting the
right duration on your bonds so that you’re not exposed to
that interest rate risk.
A bucket strategy requires
people to have a dynamic
asset allocation.
A A I I . C O M / J O U R N A L
2 4 A A I I J o u r n A l M A R C H 2 0 2 2
Bengen did showing that retirees should hold 75% stocks
and in no circumstances hold less than 50% stocks.
A recurring theme calls for higher stock allocations,
but that’s different from what target-date funds call for.
So, we’ve got this disconnect. The 4% rule style research
says to hold a high stock allocation throughout retirement.
Target-date funds, in contrast, are saying no, have a lower
stock allocation at retirement. Then post-retirement, if it’s
a “totarget-date fund, it just keeps you at that same low
stock allocation afterward. If it’s a “through target-date
fund, it might continue to decrease your stock allocation
as you go throughout retirement.
What we said was that you can go ahead and follow
the target-date fund approach, where you have a lower
stock allocation at retirement. But then what you do is you
increase your stock allocation again as you go throughout
retirement. It’s a risk-management technique because it
helps manage the worst-case scenario, which is having a
market downturn in the early retirement years.
So, a rising equity glide path—having a lower stock
allocation at retirement and then increasing that later—is
going to work best for you
when a bad market envi-
ronment occurs early in
retirement followed by a
better market environment
later on. Now, if you had a
bad market environment
for your entire retirement,
nothing’s going to work.
But a more probable worst-
case scenario is getting a
downturn early on, and
then markets recover. And
that’s where the strategy is designed to give you risk man-
agement to help you get through the sequence of returns
risk.
[See Figure 3 for an illustration of the U-shaped glide
path.]
How early before retirement should people start decreas-
ing their equities?
We never tried to optimize that side because we were
looking more at post-retirement, but I don’t necessarily
have any issues with how target-date funds approach that,
gradually over time. As you get to be maybe within 10 years
of retirement, that’s when you’re starting to get exposed
more to market returns. There is sequence of returns risk
prior to retirement as well for savers because those pre-
retirement returns are impacting their lifetime of savings.
So, that’s where you can start to take some of that risk off
the table.
I haven’t taken that analysis much further other than to
just start taking some of that risk off the table say five, 10
or 15 years before retirement.
Okay. What do we say to someone who reduced their
equity exposure heading into retirement and is nervous
about then raising their allocation? Perhaps a bear market
just ended. Any suggestions you’d give?
If you’re used to a lower stock allocation, and there was
some market volatility, it can be hard to increase your stock
allocation later. I suppose a time-segmentation strategy
like bucketing could be a behavioral way to implement
that, because you may not even realize you’re implement-
ing it. You may not realize bucketing gives you a rising
equity glide path.
If you’re coming at it from the total-return perspec-
tive, like the case study we looked at in the original article,
you’re at 30% stocks at the start of retirement and then you
work your way back up to 60% stocks.
Beyond that, just look at the historical data and see that
whenever these market volatility periods happen, things
do recover. So, just feel comfortable with that. It’s not
going to resonate with everyone. There are annuities. You
don’t have to have a total-return investment strategy for
retirement or a bucketing variation on it.
So, a rising equity glide
pathhaving a lower stock
allocation at retirement
and then increasing that
lateris going to work best
for you when a bad market
environment occurs early
in retirement followed by a
better market environment
later on.
FIGURE 3
U-Shaped Glide Path for Portfolio
Allocation
When an investor is young, most of their portfolio is allo-
cated to stocks. Then as the investor begins to near retire-
ment, exposure to stocks is reduced to protect the portfolio
against a bad sequence of returns. Once in retirement, the
investor then increases their allocation to stocks to support
spending.
Source: “Retirement Planning Guidebook,” by Wade Pfau (Retirement
Researcher Media, 2021).
A A I I J o u r n A l M A R C H 2 0 2 2 2 5
A A I I . C O M / J O U R N A L
One of our AAII asset allocation models uses a diversi-
fied version of the traditional 60% stock/40% bond alloca-
tion. For investors who are concerned about rising interest
rates, any suggestions on what to do on the bond side of the
allocation?
Everyone reading this understands the idea that if inter-
est rates go up, you can have losses on your bond funds.
And the longer the maturity or duration on your bonds,
the bigger the potential losses you experience.
If you are worried about a rising interest rate environ-
ment, that speaks more toward having lower-duration
bonds as part of your strategy. Such bonds are not going
to get as big of hit with the rising interest rates. The only
problem, of course, is that lower-duration bonds are not
yielding much at all. So, you’ve got that drag on potential
returns while you’re waiting for interest rates to rise.
Investors are in a tough spot going into retirement or
being retired in a low-interest-rate world. Even more so
in a low-interest-rate world where the expectation is that
interest rates will rise in the future. If these bonds are
meant to be the source of funds for funding their expenses
and they have to sell bonds after an interest rate increase,
then they will lock in capital losses. In this scenario, the
bonds aren’t really helping them. They’re doing the same
thing stocks do, which is creating the risk of having to sell
them at a loss.
That’s what creates the sequence of returns risk for
retirement. So, just keeping the same sort of high bond
allocation with a longer duration and so forth can create
risks if you’re worried about rising interest rates.
JOIN THE CONVERSATION ONLINE
Visit AAII.com/journal to comment on this article.
MORE AT AAII.COM/JOURNAL
Insights on Using the 4% Withdrawal Rule From Its
Creator an interview with William Bengen, January 2018
How to Score a Retirement Home Run by Paul
Merriman, September 2021
Level3 Withdrawal Strategy: S&P 500 Near High in
December 2021 by Charles Rotblut, CFA, January 2022
Previous articles by Wade Pfau and Michael Kitces on
Retirement Withdrawals:
Increasing Retirement Withdrawal Rates Through
Asset Allocation, April 2015
Reduce Stock Exposure in Retirement, or Gradually
Increase It?, April 2014