Note to Reviewer: Please arrange for posting of this article at 7:00 a.m. on Tuesday, July 25, 2017.
Valuation-Informed Indexing #351
by Rob Bennett
Michael Mauboussin: Here’s what active managers can do
The Buy-and-Hold retirement studies get the numbers wrong.
I have been saying that since the morning of May 13, 2002, when I advanced the remarkable claim to the Retire Early discussion board at the Motley Fool site. Thousands of my fellow community members at the board used the infamous “4 percent rule” to determine when to hand in resignations to high-paying corporate jobs and begin their early retirements. Because of the importance of the issue to aspiring early retirees, threads about how to use the rule appeared at that board on a daily basis. I had known about the errors in the studies for several years before I worked up the courage to point them out to the thousands of people staking their futures on their accuracy. I knew that I was going to get some flak. And it turned out that I saw about 500 times the amount of flak that I expected. That turned out to be one highly controversial post!
As the years have passed I have gradually come to a fuller understanding of why.
People still use those studies to this day. They have not been corrected. But numerous experts have come out with public statements either suggesting or outright stating that aspiring retirees should not make use of the 4 percent rule when planning their retirements. The 4 percent rule posited that a retiree can withdraw 4 percent of an 80 percent stock portfolio each year to cover living expenses with virtual certainty that the portfolio will survive 30 years of retirement (permitting a retirement that commenced on a retiree’s 65th birthday to extend to his 95th birthday). We are in a twilight zone today. Buy-and-Holders no longer feel comfortable pushing the 4 percent rule. But they don’t feel comfortable disowning it either.
To openly declare the dangers of the rule would be to set off inquiries that would likely prove far-reaching indeed. The reason why the 4 percent rule does not work is that no one withdrawal rate can be safe at all times. Shiller showed in 1981 that valuations affect long-term returns. If that is so, then stock investing risk is not stable but variable. So there can no more be a single return expectation that always applies than there can be a single safe withdrawal rate that always applies. If increases in valuations cause the long-term return from that point forward to drop, returns in excess of the long-term average return (6.5 percent real) are not even good news. All that we are doing when we push valuation levels above the fair-price level is shifting the returns on our invested money from the future into the current day; we pay a heavy price for all bull markets that we create.
So getting the safe withdrawal rate right is a big deal. Over the years I have worked with some smart people who did the math and reported that the safe withdrawal rate is a number that ranges from 1.6 percent when stock valuations are at shy-high highs to 9.0 percent when stock valuations are at rock-bottom lows. It of course follows that our realistic long-term return expectation must change with changes in valuations as well. And, sure enough, the historical return data reveals that the most likely 10-year annualized return on U.S. stocks was 15 percent in 1982 but only a negative 1 percent real in 2000. Valuations matter!
The 4 percent rule is dangerously misleading. But the mistake behind the development of the rule was not at all a random one. It was a mistake rooted in a fundamental misunderstanding of how stock investing works. The 4 percent rule follows from the belief that the market is efficient, that stock prices play out in the form of a random walk. If those things were so, the 4 percent rule really would work. The significance of the rule is that it reveals how dangerous it is to develop investing strategies rooted in fundamental misunderstandings of how the market works. The reality is that market prices have never once in U.S. history played out in the form of a random walk in the long run. If valuations affect long-term returns, long-term prices must be highly predictable.
The 4 percent rule is not the product of mere foolishness. It is the product of tough logic being applied to a fundamental misunderstanding. So the 4 percent rule possesses a certain significance even if it does not accurately identify the safe withdrawal rate for most retirements. The 4 percent rule points us to the average safe withdrawal rate, the withdrawal rate that is more or less at the mid-point of the safe withdrawal rate that applies at times of super high valuations and the one that applies at times of super low valuations.
The mid-point between 1.6 percent and 9.0 percent is 5.3 percent. So the 4 percent rule does not identify the mid-point precisely. But that’s probably just because the way in which the calculation is done is highly sensitive to a small number of bad outcomes. There is only one return pattern in U.S. history that caused a withdrawal rate of slightly more than 4 percent to fail. The Buy-and-Hold retirement studies are highly conservative studies. A withdrawal rate of 3.3 percent would work if the return on stocks for 30 years running was zero. These studies are stating that a retiree seeking safety can only take a withdrawal of 0.7 percent more than the withdrawal that would work if stocks provided a return of zero. That’s quite a claim!
So my guess is that the mid-point safe withdrawal rate is a number somewhere in the neighborhood of 5.3 percent. While the 4 percent number has in recent years been a number markedly on the high side, in most circumstances it is too low. We should not be letting recent history influence us too much. There has never before in the history of the U.S. market been a time when stock valuations have remained so dangerously high for so long a time.
The thing that is hard to accept is that even the super-conservative 4 percent number has not been low enough for retirements beginning in recent years. The zero-return safe withdrawal rate is 3.3 percent, but the real-world safe withdrawal rate dropped to 1.6 percent in 2000. The safe withdrawal rate for Treasury Inflation-Protected Securities, a risk-free asset class, was 5.8 percent when they were first issued and that was the time when the safe withdrawal rate for stocks was only 1.6 percent. Can such things be possible?
It’s because the numbers are so crazy that I think they are so important. All of the claims that I am making here follow from Shiller’s “revolutionary” (his word) discovery that valuations affect long-term returns. The crazy numbers help me appreciate just how revolutionary Shiller’s Nobel-prize-winning research was. Learning that stock investing risk is not static but variable changes everything. Thinking through why we got safe withdrawal rates wrong and why it has taken so long for us to change how we think about stock investing enough to get them right helps make the realities click for me.
Rob’s bio is here.