Journal of Financial Planning: October 2012
Jonathan Guyton, CFP®, is principal of Cornerstone Wealth Advisors Inc., a holistic financial planning and wealth management firm in Edina, Minnesota. He is a researcher, mentor, author, and frequent national speaker on retirement planning and asset distribution strategies, and a former winner of the Journal of Financial Planning’s Call for Papers competition.
It’s a financial nightmare of nearly every retiree—not to mention their financial planner—that, despite their careful planning and adherence to the tenets of safe withdrawal rate research, their retirement date will turn out to be that “perfect storm” distribution period that ultimately runs out of money. And it has been suggested that by following a static withdrawal strategy with annual inflation adjustments and rebalancing, the millennial retirement date of January 2000 could ultimately be just that scenario.
We already know that the case of the millennial retirement contains critical (and perhaps harrowing) advisory situations that caused a financial planner to face several “is it different this time” moments since distributions began in 2000, based on implementing a static inflation-adjusted distribution strategy with a 4.5 percent initial withdrawal rate using the January 2000 portfolio value.
This is based on an annually rebalanced 60/40 (equity/bond) portfolio with fixed income measured by the Barclays Intermediate Government Index (commonly used in safe withdrawal research) and equities split 2/3 Russell Global Index and 1/3 S&P 500 to provide exposure to U.S. large-cap, U.S. small-cap, foreign developed market and emerging market stocks with about 65 percent of these equities in U.S. holdings. This portfolio returned an average of 4.2 percent annually with the worst year in 2008 (–20.3 percent) and the best in 2003 (21.9 percent).
By just the fourth year, the withdrawal rate would have risen from 4.5 percent to 6.5 percent, a 44 percent increase. What to do? To what degree? And on what basis? And by 2009, the 10th year of distributions, the withdrawal rate hit 7.8 percent—a full 73 percent above its initial level. Again, what to do when the time-tested response is to stick with the plan “because research shows it’s always worked before”?
More significantly, if such moments of understandable stress convinced an adviser that they first needed to make a macro prediction (“Is it different this time?”) before they could render this advisory judgment, the chance of error, regret, or unintended consequences increased markedly, likely putting volatility in their advice (“Yes, you were fine a year ago, but now you need to significantly reduce your spending”). Indeed, for retirees to have followed advice to reduce their withdrawal rate to no more than 3 percent–4 percent of these depressed end-of-2008 portfolio values, up to a 50 percent reduction in pre-tax portfolio income (though not total income) could have been required.
What may seem like a choice between staying the static course through a near-perfect storm and changing course for empirically uncharted waters strikes me as a false choice.
Of course, it’s far too early to know whether our millennial retiree, despite following empirically based safe withdrawal principles, will ultimately exhaust their nest egg before a 30-year (or longer) distribution period ends. However, we are wise to ask this and do even better to wonder whether what we’ve learned as a profession since then would have made any difference.
My intention is to take a shot at doing just that.
A History Lesson for Withdrawal Rates
Historical analyses must be careful not to place undue advantage on hindsight. That said, I believe that revisiting the millennial retirement and examining the combined impact of key safe withdrawal research published since then can be both illuminating and build confidence.
The story of the Year 2000 retiree actually begins several years before. Imagine that it’s four years prior—very late in 1995—as this not-yet-retired client concludes a review meeting with their financial planner. Based on four years’ more work, yearly 401(k) deposits totaling $20,000, and a retirement nest egg of $664,000 today, their planner projects assets of $1 million at retirement for their “core” nest egg, given a balanced portfolio of 60 percent stocks and 40 percent bonds earning 8 percent annually. At the desired level of retirement spending and with a “safe withdrawal” income stream of $45,000 from these assets combined with projected Social Security benefits, the goal to retire in four years looks solid from a financial perspective.
But that’s not what happened. Four years later, although $80,000 had indeed been added, their nest egg’s value was not nearly $1 million. Actually, the projection missed by 23 percent. Instead, using the 60/40 portfolio as described above, these clients entered their millennial retirement with investment assets of $1.23 million!
What a problem to have: a 23 percent surplus. But as described earlier, a problem—or at least a situation calling for proper diagnosis—it nonetheless was based on research by Michael Kitces in 2008. He demonstrated that, for portfolios allocated 50 percent–60 percent to equities, distribution periods beginning in over-valued equity markets have always been sustainable when begun at a 4.5 percent withdrawal rate. And by virtually any valuation indicator—P/E, P/E10 or a shift from rising to falling real earnings—markets were indeed overvalued, with some measures putting the 2000 valuation in the highest 1 percent in over a century.
These extreme conditions, combined with the amount of additional assets available to meet the client’s spending goal, make it prudent to recommend withdrawals at a 4 percent rate, rather than at 4.5 percent. Assuming the millennial retiree’s spending goals did not rise during these four years with their good fortune in returns, this adjustment compromises nothing. With assets over 20 percent higher, an initial withdrawal rate 10 percent lower can readily generate the income called for in their plan.
Even by itself, this would have made a difference. By 2009, the withdrawal rate would have been 6.4 percent and not 7.8 percent. Admittedly, this would still have been a serious situation, but the impact of this single advisory judgment is noteworthy. Happily, other improvements come from the application of recent research.
Dynamic Allocation and Withdrawal
In 2009, Kitces demonstrated the benefit of incorporating a dynamic asset allocation approach in the face of over-valued and under-valued markets (as did Wade Pfau in 2012). Between them, they explored shifts of 33 percent–50 percent. For example, in a market deemed over-valued, a 60 percent valuation-neutral target equity allocation would be reduced to 40 percent via a 33 percent reduction and to 30 percent with a 50 percent shift.
I utilized a dynamic allocation policy for the millennial retiree that reduces their equity allocation by 17 percent (from 60 percent to 50 percent) in over-valued markets. This occurred from 2000 until 2009, when markets again became fairly valued for that year only. At no point was the equity allocation increased above the 60 percent target with a dynamic policy shift, because that valuation threshold was never crossed; however, some advisers may have done so at least once.
Compared to the static 60/40 portfolio’s 4.2 percent average annual return, the dynamic allocation policy return was 17 percent higher (4.8 percent)—with a best year of 20.7 percent and a worst of –15.2 percent. Interestingly, the 10.2 annual standard deviation was 17 percent lower than the 60/40 portfolio. Implementing this policy from the inception of distributions produced a 48 percent increase in the withdrawal rate (to 5.9 percent) by 2009, much improved though still disturbing.
The benefits of a dynamic withdrawal policy were demonstrated in research by Guyton and Klinger in 2006. We found that a withdrawal policy freezing the distribution amount after years with an investment loss and lowering the real distribution amount by 10 percent, when the withdrawal rate would have otherwise risen by 20 percent, generated a 1.0 percent rise in the safe withdrawal rate. (Later that year, Bengen showed that a smaller adjustment trigger produced a half percent withdrawal rate increase.)
Normally, this would have justified a 5.5 percent initial withdrawal rate. However, given the extreme market over-valuation in 2000 (both P/E measures were easily in the highest 3 percent of readings since 1880), I implemented these policies using a 5.0 percent initial withdrawal rate.
Significantly, because this dynamic withdrawal policy begins distributions at a withdrawal rate (5.0 percent) that is 25 percent higher than with the static approach (4.0 percent), the core portfolio supporting these distributions need only be 80 percent as large to produce the same income. These remaining retirement assets may be seen as either a discretionary fund for additional non-recurring expenses (bucket list?), a super-sized emergency fund (should following the dynamic withdrawal policies lead to a decline in real income), or some blend of the two. The higher asset level required to generate the same income under the static approach often makes it impossible to create such a fund.
The Client-Ease (and Planner-Ease) Factor
At our firm, this concept has increased retirees’ peace of mind immeasurably—especially among couples. We have found it important, though, to present any such surplus as a finite lump sum (which is only replenished via below-policy spending or new assets, such as an inheritance) rather than as a higher income that easily can be absorbed with higher ongoing expenses.
Applying these dynamic withdrawal and allocation policies in combination dramatically changed the millennial retiree’s situation. Withdrawal rates were far less volatile throughout the distribution period. By the start of 2003 it had indeed risen to 6.1 percent, but by 22 percent rather than 33 percent with dynamic allocation policies alone and 43 percent with none. As 2009 began in the throes of financial crisis, it had climbed to 6.4 percent (28 percent higher than in 2000), but far below that of the other approaches—the most stable of which rose by 48 percent by then.
Without question, this additional stability in withdrawal rates during this time was accomplished by the two 10 percent nominal cuts and one freeze the dynamic withdrawal policies triggered. Absent any corresponding real expense reductions (mortgage refinance, voluntary frugalness in uncertain times), using about 20 percent of the original discretionary fund could make this up.
Unexpectedly, this analysis also suggests these policies can at least somewhat mitigate the impact from sequence-of-returns risk. Though clearly an unfavorable investing period, reversing the order of returns (with gains in seven of the first nine years and just one equity downdraft rather than gains in six of nine with two bear markets) generates a portfolio value 9 percent higher when no dynamic policies are implemented. No surprise there, as the reverse order is more favorable when ongoing distributions occur. However, with dynamic policies in force, the same reversal created one-third less difference—an encouraging defense.
Clearly, the use of recent empirically based dynamic allocation and withdrawal policies have had a significant and stabilizing impact on the challenging case of the millennial retiree—at least thus far. By 2012 their withdrawal rate was 5.5 percent, 10 percent above its 2000 level compared with a 55 percent leap had no such policies been implemented.
These material financial improvements notwithstanding, perhaps more important is the reduced advisory stress that otherwise could have caused both untold—and unnecessary—volatility in both portfolio distributions and advice under a static approach. While there is undoubtedly much more to learn about improving the sustainability of portfolio withdrawals as well as our ability to serve clients, we need look back no further than the past dozen years to see an encouraging financial, emotional, and psychological demonstration of our profession’s collective advancement. Let’s hope it remains sustainable!