The Staggering Impact of Sequence of Returns in Retirement Portfolios

Jul 6, 2017 / By Craig L. Israelsen, PhD
Print AAA
Add to My Archive
My Folder

My Notes
Save
It’s not uncommon to compare mutual funds on the basis of their long-term return. But beware. These funds have identical 47-year performance, but shockingly different results when used as the sole ingredient in a retirement portfolio subject to annual RMDs. The key factor: Sequence of returns drastically influences growth in rolling 25-year periods between 1970 and 2016.

Editor’s note: For over 10 years, Horsesmouth has been bringing you, our readers, the Top 10 most popular articles of the year during the holiday season. In 2017, advisors were keenly interested in the sweeping changes affecting the industry—and adapting to those changes for continued success. Watch this space for this year’s best ideas on developing business in the 21st century, maneuvering in the new technological environment, and the continued primacy of maintaining deeply personal client service and relationships. From more than 500 articles published in 2017, Horsesmouth members rated this as one of the best.

It’s not uncommon to compare mutual funds on the basis of their long-term return. In this case, how about over the past 47 years?

Interestingly, there are only 107 mutual funds that were in existence in 1970 and are still alive today (as of December 31, 2016). That tally is based on mutual funds in the STEELE Mutual Fund Expert software. The list of these 107 long-standing funds is shown below in Table 1.

From this list, I will be comparing two sets of funds that had identical 47-year performance (as measured by a single lump sum investment), but demonstrated very different behavior when used as the sole ingredient in a retirement portfolio subjected to annual RMD-based withdrawals. The key differentiator: sequence of returns. As you will see, sequence of returns makes a huge difference in how the retirement portfolio fared during 23 rolling 25-year periods between 1970 and 2016.

Table 1: Mutual Funds With a 47-Year Performance History (1970-2016)

Source: Craig Israelsen, PhD

The first set of funds to be compared are two venerable veterans in the mutual fund universe: Franklin Mutual Shares Z and Fidelity Magellan. Both had a 47-year average annualized return of 12.94% between January 1, 1970 and December 31, 2016 (based on a single lump-sum investment with no additional investments and no withdrawals). However, when used as the sole investment in a retirement portfolio, the outcomes were vastly different in many cases.

Each fund started with $1,000,000. Annual withdrawals were determined by the required minimum distribution (RMD) over 23 rolling 25-year periods from 1970-2016. The first 25-year period was from 1970-1994, the second was from 1971-1995, and so on. This approach assumes the retiree went from age 70 to age 95. Only the amount of money specified by the RMD was withdrawn—no more. The performance of each fund was measured across three variables: (1) ending account value after 25 years, (2) average annual RMD withdrawal during each 25-year period, and (3) total amount of money withdrawn over each 25-year period.

As shown in Table 2, below, these two funds often had dramatically different performance when being used as a retirement portfolio. For example, let’s consider the period from 1975-1999—a period with impressive performance for both funds. Franklin Mutual Shares had a 25-year annualized return of 19.39%, whereas Fidelity Magellan had a return of 25.88% (lump-sum based “accumulation-mode” performance).

Table 2: Franklin Mutual Shares and Fidelity Magellan Retirement portfolio with a $1,000,000 starting balance Annual withdrawals determined by the required minimum distribution

Source: Craig Israelsen, PhD

However, the difference in each fund’s performance when employed as the sole investment in a retirement portfolio was almost beyond belief during this particular 25-year period (highlighted with yellow). Franklin Mutual Shares had an ending balance of $20,696,923 whereas Fidelity Magellan had an ending balance of $82,852,774 when the retiree was 95 years old. Both started with $1,000,000 when the retiree was 70 years old.

A total of $17,518,493 was withdrawn by the retiree from Franklin Mutual Shares over this particular 25-year period (1975-1999), for an average annual RMD withdrawal of $700,740 (the actual amount withdrawn each year varied greatly, but $700,740 was the average). Fidelity Magellan yielded up a total withdrawal of $48,926,578 for an average annual RMD withdrawal of $1,957,063. These results are mind-blowing, but this particular 25-year period was one in which Fidelity Magellan had an ideal sequence of returns. By starting in 1975, the retiree’s portfolio avoided the U.S. stock market crash of 1973 and 1974. And by ending the 25-year period in 1999, it avoided the dot-com crash that began in 2000. Lucky retiree.

Fast forward to the most current 25-year period: 1992-2016 (highlighted in pink). The tables turn. Franklin Mutual Shares is the far better performer. Its ending balance in 2016 was $2,455,921 and the retiree was able to withdraw an average of $159,107 each year (as determined by the RMD divisors). Fidelity Magellan, on the other hand, finished with an ending balance of $1,358,767. The retiree using Magellan had an average annual RMD withdrawal of $105,245.

On the whole, Fidelity Magellan was the better retirement portfolio “engine” if measuring by the median annual withdrawal over all 23 rolling 25-year periods. Franklin Mutual Shares had a median annual withdrawal of $352,759 compared to $479,480 for Fidelity Magellan. Admittedly, the annual withdrawals in the more recent 25-year rolling periods were lower than the overall median figure over the past 47 years.

There are several key observations:

  1. Sequence of returns had no impact on the overall 47-year, lump-sum, average annualized return, but manifests itself dramatically when evaluating a fund that it is being subjected to annual withdrawals (as in the case of a retirement portfolio).
  2. The different results across the various 25-year periods is astonishing. A retiree using Fidelity Magellan as the only fund in their retirement portfolio had a positive “perfect storm” experience during the period from 1975-1999. During that period the retiree withdrew an average of nearly $2 million each year from their portfolio and still ended up with a balance of nearly $83 million at the age of 95 (again, if only withdrawing the amount specified by the RMD). A different retiree, who retired at age 70 in 1992 had a very different experience using Fidelity Magellan as their retirement portfolio. Their average annual RMD withdrawal over the next 25 years was just north of $105,000 and their portfolio balance at age 95 (at the end of 2016) was just over $1.3 million. Certainly not a bad outcome—but very different from the retiree who used Magellan between 1975 and 1999. The difference? Sequence of returns.

Rely on luck?

How do retirees protect themselves against the randomness of sequence of returns in their retirement portfolio? The most fundamental approach is to build a diversified retirement portfolio that is not subject to just one asset class. Fidelity Magellan and Franklin Mutual Shares are both equity funds (with Mutual Shares also having some exposure to non-U.S. stocks and bonds/cash at times).

The key to genuine diversification, and therefore more protection from an adverse sequence of returns, is allocating the retirement portfolio across a wider variety of asset classes. Disparate asset classes, by their nature, tend to have different performance patterns. Thus they create a variety of return sequences that ideally blend into a smoother path of performance. In short, the highs and lows of performance are flattened out.

To illustrate the smoothing effect produced by utilizing a diversified fund in a retirement portfolio, I will compare Eaton Vance Balanced A (a balanced fund that has roughly 60% stock and 40% bonds) against VanEck International Investors Gold A (a sector fund that focuses on precious metals companies—the antithesis of being broadly diversified). Both funds had identical 47-year lump-sum annualized returns of 8.85% from 1970-2016—hence they were chosen for comparison.

As shown in Table 3, below, the ending account value and average annual RMD withdrawal for Eaton Vance Balanced (which is used as a representative for “balanced funds” in general) was relatively stable across the various 25-year rolling periods. We don’t see the stratospheric results during the 1975-1999 period that we observed in Mutual Shares and Magellan. Rather, a diversified fund (at least, diversified across two asset classes) dampens the variability across the different 25-year periods or, in other words, reduces the impact of sequence-of-returns risk. We don’t hope that our sequence of returns is lucky. Rather, we diversify so as to insulate ourselves from a less than favorable sequence of returns in one particular asset class (such as U.S. stock).

Table 3: Eaton Vance Balanced A and VanEck International Investors Gold A Retirement portfolio with a $1,000,000 starting balance Annual withdrawals determined by the required minimum distribution

Source: Craig Israelsen, PhD

Naturally, diversifying also reduces the upside potential, meaning that a balanced fund won’t have the impressive results (massive ending account value and huge annual RMD withdrawals) that a less diversified fund can experience during a period of time with an ideal sequence of returns. As a side note, the balanced funds with the best performance (both in terms of lump-sum return and performance when used in a retirement portfolio) were Dodge & Cox Balanced, Franklin Income A, and Mairs & Powers Balanced). Each of those three funds had a median ending account balance of approximately $3.5 million and an average annual RMD withdrawal of around $180,000.

Using a sector fund, such as VanEck International Investors Gold A (which is a representative of many types of sector funds), as the sole fund in a retirement portfolio is truly rolling the dice. In several of the 25-year periods it worked out great, such as 1970-1994 or 1972-1996. But in other 25-year periods it was a bust, such as 1981-2005, 1984-2008, 1989-2013, etc. This illustrates that a non-diversified fund (such as a sector fund) is very susceptible to sequence-of-returns risk. And, as illustrated, the risk was not worth it. The median ending account balance for the VanEck sector fund was $985,667—far lower than the $2,484,812 achieved by Eaton Vance Balanced. The average annual RMD withdrawal for VanEck Gold was $70,270, versus $135,147 for Eaton Vance Balanced.

Bottom line: If you want to mitigate the potentially negative impact of a bad sequence of returns in your client’s retirement portfolio diversify their portfolio more broadly.

Craig L. Israelsen, PhD, is an Executive-in-Residence in the Personal Financial Planning program in the Woodbury School of Business at Utah Valley University. He is the developer of the 7Twelve® Portfolio. Learn more at www.7TwelvePortfolio.com.

Comments

Amazing!!! Great article!!! Seems like a good answer to the DIY-do it yourself retiree. Thank you.
Great article. Thanks.
Nice presentation of the impact of "sequence of return". While I think - long term - the unfortunate issue with this article is the time frame has fairly good starting dates whether it is 1975 or 1991. Are you able to get Craig to show his analysis using a start date of 2000? While not a 25 year retirement picture i believe we would all benefit from seeing a very different (negative) impact of sequence of return. Mike
This is a great illustration of sequence of returns and the impact of diversification. What would the results look like if the draw down was 4% instead of RMD?

IMPORTANT NOTICE
This material is provided exclusively for use by Horsesmouth members and is subject to Horsesmouth Terms & Conditions and applicable copyright laws. Unauthorized use, reproduction or distribution of this material is a violation of federal law and punishable by civil and criminal penalty. This material is furnished “as is” without warranty of any kind. Its accuracy and completeness is not guaranteed and all warranties express or implied are hereby excluded.

© 2024 Horsesmouth, LLC. All Rights Reserved.