The mistake referenced in the title is a misunderstanding of withdrawal rate.
I was recently speaking with a successful entrepreneur about sustainable withdrawal rates from investment funds, particularly retirement funds.
He said that he had heard that 4.5% was a “sustainable” withdrawal rate, and that anything higher was too risky.
Yet, we recently conducted a simulation of a tax-exempt trust that was invested in the US stock market and paying out 5% every year.
We looked at 77 rolling 50 year periods, beginning in 1945 and ending in 2022. We assumed that each trust began with $1,000,000, and distributed 5% of its beginning year value every year for 50 years. We then took the amount of the trust balance in year 50, and put that in our histogram, below.
- The definition of a withdrawal rate
- The expectation that accounts will be invested in a balanced portfolio
Definition of Withdrawal Rate
The withdrawal rate question usually arises in the context of retirement planning. Suppose a client retires with $1 million in a retirement account. The “sustainable withdrawal” rate question is actually a question of sustainable dollar amount, not percentage rate.
When people[2] say that a sustainable withdrawal rate is 4.5%, often (as is the case with the Fidelity example cited) what they really mean is that a fixed dollar amount is to be withdrawn. In the first year, that fixed dollar amount is $45,000 (on a million). That is 4.5% of the initial $1 million, but the withdrawal amount in subsequent years will likely be different from 4.5% of the amount in the retirement account.
In fact, in the example Fidelity offers, they suggest[3] that the amount be adjusted by inflation. In the modern world, that means it is very probable that the amount will increase each year.
Balanced Portfolios
Balanced portfolios are a second reason that a fixed and growing 4.5% of the initial retirement balance may expose the retiree to an unacceptable risk of running out of money.
The reason is that, although balanced portfolios (e.g. 60% equities, 40% bonds) are generally less risky than all equity portfolios (i.e., they have less risk of decreasing sharply in value), they also have significantly lower expected returns.
This lower expected return is a mathematical fact, as long as bonds have a lower expected return than equities. Suppose, for example, that we expect the long run rate of return on equities will be 10% compounded annually,[4] and that the long run expected return on bonds is 5%[5].
The math is simple. Sixty percent of 10% is 6%, and 40% of 5% is 2%, giving an expected total return on the 60/40 portfolio of 8%.
Variability
An 8% long run average return would be pretty good, and would seem to be more than enough to offset annual distributions based on 4.5% of the starting value.
But many simulation studies suggest that even at 4.5% of the initial amount, the distribution might not be sustainable.
The reason is variability of returns. Some advisors call this “path dependency.”
The math is a little involved, but the concept is easy.
Suppose a portfolio begins with $1,000,000, and the withdrawal amount is set at $45,000, to be adjusted each year by inflation.
Now suppose that we get a year like 2022. In 2022, US stocks returned -18%, and US bonds had an equally dreadful year.[6] And to make matters worse, inflation, as measured by the CPI, was 6.5%.
Let’s look at what happens with our hypothetical portfolio. It starts at $1 million. The market knocks 18% off.[7] The portfolio is now down to $820,000. We also have to withdraw the $45,000, leaving an amount remaining of $775,000.
If the withdrawal rate really were 4.5%, the withdrawal would be sustainable.
But the withdrawal rate is not 4.5%. It is $45,000, and then we increase that by inflation of 6.5%, so the withdrawal in year two is $47,925.
That second year withdrawal amount of $47,925 represents 6.2% of the remaining portfolio value of $775,000.
A single such year is not likely to be a problem. But two or three bad years near the beginning of a withdrawal plan could easily derail it, particularly if those bad years are accompanied by inflation.
True Percentage Withdrawals
One mathematically effective solution is to base withdrawals on a fixed percentage of the account value each year. Done this way, it is mathematically nearly[8] impossible for the account to run out of money.
The simulation shown at the beginning of this post offers two important lessons. One is the lesson that there is an important difference between a fixed withdrawal amount often mistakenly considered as a percentage, and a true percentage that is the same percentage every year. The second lesson is that over long periods, such as 50 years, the main effect of a balanced portfolio has been to depress long run returns.
Long-Term Tax Exempt Trusts
It is possible to create a tax-exempt trust for clients who have long term investment goals. For more information, click here to request free copy of our Advisor Guide to Stock Diversification Trusts. Or email your request to [email protected], or call us at (703) 437-9720 and ask for Connor or Katherine.
[1] The purpose here is not to review the studies, but one such study is An International Perspective on Safe Withdrawal Rates from Retirement Savings: The Demise of the 4 Percent Rule? By Wade D. Pfau, GRIPS Research Center, Discussion Paper 10-12.
[2] For example, Fidelity suggest “aim to withdraw no more than 4% to 5% of your savings in the first year of retirement, then adjust that amount every year for inflation.”
[3] https://www.fidelity.com/viewpoints/retirement/how-long-will-savings-last#:~:text=The%20sustainable%20withdrawal%20rate%20is,amount%20every%20year%20for%20inflation.
[4] The pre-tax, dividends-reinvested compound return on the S&P 500 has been about 9.9% since 1928. That includes terrible periods of the 1930s, the 1970s, and the 2007-2011 financial crisis.
[5] This is closer to the yield, rather than the total return, to be expected from corporate bonds because it makes no allowance for defaults, but that is a technicality that does not affect the overall point.
[6] https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/histretSP.html
[7] The timing of the withdrawals during the year matters, but not decisively for the purposes of this illustration.
[8] We say “nearly” impossible because in the real world, frictions can disrupt the mathematical fact that a constant percentage withdrawal rate will, at worst, cause the account balance to approach, but never reach, zero.
Leave a Reply