June 24, 2021
One of the unanticipated effects of the pandemic is that it
has induced millions of workers to retire early. John, for
example, who is now 62, had previously expected to work
until 65 or 68. He has $1 million of financial assets, which
makes him better off than many of the involuntary retirees,
and the investment advisor he consulted suggested that he
place 30% of his assets in common stock and 70% in
interest-bearing securities. Yet the advisor cannot answer
his major question: how much spendable funds can he draw
safely, without the fear of running out of assets while he
is still alive?
The Decision Process
Decisions bearing on this question should be informed. An
informed decision is one based on a known or estimated
probability of outcomes. If the probabilities are not known,
the decision is uninformed.
The decisions by retirees regarding how much they can safely
draw is largely uninformed. This is the case whether the
retiree follows her own instincts or relies on a financial
advisor for an answer. Yet to a very large extent, these
uninformed decisions can be converted into informed
decisions. How to do that is the major purpose of this
Spendable Fund Availability During Retirement
How much John can safely draw depends on how long he lives
and the rate of return on his assets. While neither is
known, both can be expressed as probabilities. Mortality
statistics show the likelihood that John will still be alive
at various ages. For example, he has a 53% probability of
reaching 82 but only a 15% probability of reaching 92. These
probabilities are shown in the top section of the attached
Also known is the estimated probability of different rates
of return, over periods of varying length, based on the past
history of rates of return on similar portfolios. This is
shown in the lower section of the table. For any given
period, the distribution of rates with their probabilities
also shows the distribution of draw amounts having the same
probabilities. The draw amounts shown are the initial
amounts subject to an inflation adjustment of 2% a year.
A Simple Example
Suppose John’s cardiologist tells him that he will surely
die in 5 years or less. That bad news would at least
simplify his decision regarding how much of his assets he
could safely draw each month. His decision process could be
based entirely on the rate of return on his portfolio over 5
As shown in the table, the median rate of return over a
large number of 5-year periods was 6.90%. If that rate
materialized, he could draw $18,934 each month for 60 months
and die penniless.
Drawing the median amount would be imprudent, however,
because in half of the 5-year periods, the rate of return is
less than 6.9%. Indeed, there is a 2% probability that the
rate will be negative (-0.55%). If John drew a payment based
on a 6.9% rate and the realized rate turned out negative, he
would become penniless before he died. It is plausible,
therefore, that John would avoid that risk by drawing less —
$15,785 would last for the full 5 years even if the rate of
return was as low as -.55%.
A More Plausible Case
Suppose that John’s cardiologist emails him with surprising
news: he mixed up the medical charts and John’s health is
fine — he might live to 92 in fact.
The flip side of this good news is that the longer he
expects to live, the less he can safely draw. If he assumes
that he will live 30 more years to age 92, the median return
of 6.41% over 30-year periods would support an initial draw
of $4965 a month.
In this case John would face 2 risks. One risk, as explained
earlier, is that John’s realized return will be less than
the median return. If he draws $4965 initially and the
realized rate turns out to be less than 6.41%, his assets
will run out before he hits 92.
The second risk is that he will live beyond age 92. The
probability that this will happen is about 15%. Even if the
median rate of return of 6.41% and the associated initial
payment of $4,965 are realized, John’s assets and monthly
payment will drop to zero when he hits 93.
Adjustments Over Time
The analysis to this point has focused on how John’s initial
retirement plan plays out over time. The reality is that as
the realized rate of return diverges from the plan, and as
he ages, he could adjust the monthly draw amount as needed,
perhaps in consultation with a financial advisor. But
retirees may not want to incur the expense of an investment
advisor, and few want to spend their retirement years
managing their portfolio. A needed reduction in the payment
is troublesome, furthermore, and may be resisted. There is
another way that many retirees might choose if offered the
An Alternative Approach:
the Asset/Annuity Combo
The alternative reduces rate of return risk, eliminates
mortality risk, and minimizes the need for surveillance.
This alternative is a combination of asset management and
annuity, and it is illustrated by the rightmost column of
An asset-annuity combo would use John’s assets in two ways.
Part of his assets is used to purchase a deferred annuity,
which is paid for upfront but doesn’t start paying the
retiree until a specified deferment period is over. In the
case shown in the table, the deferment period is 10 years.
Starting year 11, John will begin receiving annuity payments
for life, which eliminates the risk of living too long.
The remainder of John’s assets are used to provide spendable
funds during the 10-year deferment period. He is subject to
rate of return risk during this period, but it is much
reduced and more manageable than being dependent on asset
draws over his remaining life. This is reflected in its
smaller variance in initial draw amounts over different
rates of return.
The technology used to deliver the combo, developed by my
colleague Allan Redstone, provides a seamless transition
from asset-draws to annuities, along with a pre-specified
inflation adjustment. It incorporates HECM reverse mortgages
where it is relevant for a retiree.
Providing Estate Value
Retirees who place high value on leaving funds to an estate
can easily modify a combo plan by segregating a part of
their assets which would not be used in determining the
amount of spendable fund draws. These “set-aside” assets
targeted to the estate could be made available only in the
event that the rate of return falls short of the rate
assumed in calculating draw amounts.
The combo is not needed by retirees with sufficient wealth
that their retirement spending would not be affected either
by unexpected changes in rates of return on financial assets
or by living well past their expected life. The target is
the much larger group of retirees like John, who have
limited wealth and want to draw the largest possible flow of
spendable funds from it with no risk of ever running out.