A couple of weeks ago, a respected and rather mathematical member of the popular Bogleheads online investing forum, Siamond, posted a thoughtful article, CAPE and Safe Withdrawal Rates, at the Bogleheads blog.
In the post, and updating analysis done by retirement researcher Michael Kitces a decade earlier, Siamond looks at whether stock market valuations impact the withdrawal rate a retiree can safely use.
The Worry About Withdrawal Rates
Just to put us all on the same page, and keeping the math simple, say someone has $1,000,000 as their retirement nest egg. (Most people, by the way, don’t have anywhere near this amount. I use $1,000,000 to keep the math easy.)
Popular thinking says you or I could, with $1,000,000 of retirement savings, draw $40,000 annually. Maybe even a bit more. And that that withdrawal rate will almost surely last 30 years.
Further, with a plan like this, you and I can give ourselves annual “cost of living” raises for inflation. If inflation runs 2%, we can bump $40,000 to $40,800.
This all sounds pretty good. But with the stock market richly, richly valued, people near or in retirement prudently ask a scary question: Do current high stock market valuations change the math?
In other words, do you and I need to use a lower withdrawal rate if the stocks that make up most of our retirement savings trade at 30 times earnings rather than, for example, 15 times earnings (the average).
Siamond discusses this worry in detail, and he concludes the data do not paint a “pretty picture.” High stock market valuations probably do lower the rate you or I should spend down our nest eggs. At least if we want to be cautious.
Note: Siamond uses the ratio of the stock price to the ten-year average earnings, more popularly known as the “cyclically adjusted price earnings” ratio and also as the CAPE 10 ratio, to quantify high stock market valuations.
I think Siamond is right. And that those of us in retirement and close to retirement need to consider his work if we’ve been using more optimistic spend rates such as the well-known 4% “safe withdrawal rate.”
Accordingly, this post reviews the Siamond withdrawal rate math.
And then it makes a suggestion about the relatively easy “fix” some of us can use to repair our broken retirement plans.
Siamond Withdrawal Rate Math
The Siamond withdrawal rate math works like this: You estimate the stock market return you’ll earn over your retirement by dividing 1 by the CAPE ratio.
Right now, for example, the CAPE ratio for US stocks equals 32. If you divide 1 by 32, you get 3.1%. And that 3.1% then becomes your estimated long-run real return for the US stock market component of your savings.
Note: The research says you can’t use CAPE to estimate stock market returns over the near future, but that CAPE does a pretty good job helping you guess about what you’ll earn over ten to twenty-five years.
You then calculate a weighted average return for your portfolio by multiplying the expected returns of each slice of your portfolio by the percentage you allocate to that slice.
If you put half your money into US stocks expected to earn (say) 3% and half into US Treasury bonds expected to earn (say) 2%, the weighted average equals 2.5%.
As Siamond shows in his essay, this weighted average–2.5% using the values from the preceding sentence–becomes a pretty good though “conservative” withdrawal rate.
Sure, not a perfect safe withdrawal rate as he goes to some length to explain.
And, oh yes, still a safe withdrawal rate that you need to temper with common sense and periodic tweaks.
But all in all a pretty good and conservative withdrawal rate in that it means you almost surely don’t come up short.
Note: You can get current CAPE ratios for US stocks and international stocks from the Star Capital website. And Siamond in his essay plugs the intermediate US treasury bond rate as 2%, which seems imminently reasonable.
Dealing with the Disappointment
Okay. I’m going to guess that if you do the math, you’ll going to come up with a 3%-ish withdrawal rate.
If like me you’ve been using a 4% or 5% “safe withdrawal rate,” Siamond’s calculations suggest a reassessment of your or my retirement income and spending.
With $1,000,000, for example, you might have thought you could safely spend $40,000 a year.
The math of the Siamond withdrawal rate suggests you and I ought to use a figure more along the lines of 3%. That equals $30,000 of income if your starting nest egg equals $1,000,000.
And if you want $40,000 of income but your withdrawal equals 3%, you actually need $1,333,333 not $1,000,000 of retirement savings.
Clearly, a big change in plans. Which begs the question: What do you, what do I do?
Making the Siamond Withdrawal Rate Work
You’ve got a variety of options: saving more before you retire if you’re not yet retired, investing in assets that (hopefully) earn a higher return than US stocks, spending less after retirement, and maybe just accepting a higher potential failure rate.
All of those are good ideas. And you and I probably want to consider all of them.
But the easiest way to deal with this situation if you’re still working? I think you work a bit longer. Just working another two or three years may be all you need to adapt your original plan for the research results Siamond shares in the Bogleheads blog post.
Working longer works really well for four simple reasons:
First, every year you work, you get another year of compounding. That means you’ll start your retirement with a bigger nest egg. With a $1,000,000 portfolio and an expected 3% return, for example, every year hopefully means another $30,000 of growth in your savings.
Second, every year you work, you may be able to save an additional amount for retirement. An extra $5,000 or $10,000 for an extra two or three years makes a difference. That further pads your savings.
Third, every year you delay retirement and the point you start drawing Social Security benefits, you bump up your ultimate Social Security benefit amount by about 8%. A $2,000 a month benefit, for example, maybe grows to $2160 a month if you work an extra year, and to $2330 if you work two years. Note that bumping your Social Security benefits by $4,000 a year takes the place of about $130,000 of retirement savings if your savings earn 3%.
Fourth, if you work additional years, you may be able to amortize your retirement nest egg over fewer years of retirement. Every extra year you work may, in effect, bump your nest egg by an amount equal to the draw you would have taken had you been retired. (With the example numbers used in the preceding paragraphs, that sort of means every extra year of working adds $30,000 to $40,000 to your savings.)
Final Comments
I don’t want to sugar coat the reality suggested by high stock market valuations. If you were planning on drawing 4% or, gulp, 5% a year starting soon, Siamond’s research suggests you need to redo your arithmetic.
However, if you can finagle your way to another two or three years of work, that’ll pretty easily mitigate the withdrawal rate reality his blog post highlights.
Another comment: Other than maybe bumping up your savings, I don’t think you worry about the high stock market valuations if you’re in your twenties or thirties. Meager returns over the next decade or two probably don’t matter as much to you.
Other Resources You May Find Useful
I urge you to read Siamond’s blog post, but if you’re worried about high stock market valuations, you might find this post interesting (and stress reducing): Bear Market Survival Tactics from David Swensen.
The blog series we did on developing a Retirement Plan B might be useful, too. That series starts here: Retirement Plan B: Why You Need One.
Finally, if you’re married or have a domestic partner, can I suggest you take a peek at this blog post too? Retirement Spending and Joint Life Expectancy.
Siamond says
Very honored that you spent all this effort summarizing my research, and suggesting ways forward! I totally agree with you. I would also suggest that retirees should actively investigate variable withdrawal methods (e.g. Guyton-Klinger decision rules, PMT-based actuarial methods, etc). This is really the key to navigate good and bad times.
PS. all credits to Michael Kitces who had this remarkable insight back in 2008 about the relation between SWR and CAPE.
Steve says
Very good, Siamond. Very good. And thank you (and Mr. Kitces) for the way your work moves our knowledge forward.