November 2, 2022
In my post three weeks ago, I happily declared that the 4% Rule works again, thanks to the much more attractive equity and bond valuations. It’s always fun to deliver pleasant news. But keep in mind, everyone, that this refers to today’s retirees with their slightly depleted portfolios. But how about the folks who were unlucky enough to retire earlier this year in January 2022, when equities were at their all-time high? That cohort is off to a bad start, to put it mildly. Of course, it’s too early to tell what should have been the appropriate safe withdrawal rate for that cohort. We’re only less than a year into a multi-decade retirement. My recommendation back then would have been that due to the wildly expensive equity valuations and low bond yields one should have treaded a bit more cautiously. Maybe do 3.50-3.75% for a 30-year traditional retirement and 3.25% for a 50 or 60-year early retirement. And maybe raise that a little bit again depending on your personal circumstances, especially if you expect large supplemental cash flows from pensions and Social Security later in retirement, see my Google Simulation sheet (Part 28 of my SWR Series). Also notice also that with my estimates, I’m a bit more aggressive than the widely-cited Morningstar study recommending a 3.3% safe withdrawal rate for a 30-year retirement.
But recently, I’ve come across some rumblings that put into question all this cautious retirement planning. The reasoning goes as follows: First, the year 2000 retirement cohort actually did reasonably well with the 4% Rule. Second, the Shiller CAPE at the peak of the Dot-Com bubble was higher than in 2022. Bingo! The 4% Rule should do really well and even better for the 2022 cohort, right? I’m not so sure. That line of reasoning is flawed, for (at least) two reasons: First, the pre-Dot-Com-Crash retirement cohort experience wasn’t as pleasant as some people want to make it now. And second, I actually believe that the fundamentals in late 2021 and early 2022 were not very attractive at all. In fact, in some crucial dimensions, they were significantly worse than at the height of the Dot-Com bubble. Hence today’s post with the slightly scary and ominous title. Two days late for Halloween, I know.
Let’s take a look…
The year-2000 retirement cohort recap
A persistent myth circulates in the personal finance and retirement planning community, namely that the cohort that retired in the year 2000 at the peak of the Dot-Com bubble actually did really well since then. Some people even claim that the 2000 retirement cohort’s portfolio recovered back to its initial portfolio value. And that’s net of withdrawals! First, let’s put that to the test. Let’s examine the worst-case cohort at the Dot-Com peak:
- Initial portfolio $1,000,000, 60% S&P 500, 40% 10y benchmark bonds.
- Start withdrawing $3333.33 monthly, starting on August 31, 2000.
- Withdrawals and portfolio values are CPI-adjusted.
- I run the simulations up to October 31, 2022.
Let’s see how the portfolio value has evolved over time:
Indeed, the portfolio held up reasonably well. As of October 31, 2022, 22 years into retirement, you still have about 50% of the initial inflation-adjusted value. If you had retired back then with a 30-year horizon you would now look really solid. It’s highly unlikely that you will run out of money with only 8 years left.
But notice a few caveats:
First, at no time during this cohort’s retirement would you have even come close to recovering your initial portfolio value. This myth circulates on the web because people are ignoring inflation (again). Yes, the nominal portfolio value would have temporarily come back to above $1,000,000 in 2020, only to drop again in the most recent bear market. But you can’t have your cake and eat it too; either we keep the $40,000 in (real) withdrawals constant in which case the portfolio never recovered. Or the nominal portfolio recovered to $1,000,000 but the withdrawals were much higher than $40,000.
Even worse, at the bottom of the 2009 bear market, you would have depleted the portfolio down to about $467k. Most retirees would have likely been too scared to keep withdrawing $40,000 p.a. at that point. Not knowing how well and how swiftly the market would recover between 2009 and 2021, most retirees would have likely reduced their withdrawals at that time. And even worse, if they had thrown in the towel and gotten out of stocks at or near the bottom of the bear market in 2009, they would have never recovered. So, I always mark the 2000 cohort as one of those Trinity Study successes that were really a failure in disguise. I like to use the analogy of “Trinity Airlines” where your plane landed safely at its destination, but along the way, the engines were on fire and the pilot was screaming “we’re all going to die!”, see item #5 in Part 26 of the SWR Series). Not really a success!
And also keep in mind that the 60/40 portfolio allocation was the best possible case. If you indeed had lacked the foresight to put your money into that exact allocation, your August 2000 retirement could have been worse. For example, if you think back to the height of the Dot-Com bubble when everyone was drunk with confidence that the stock market will grow at double-digit rates forever you might have felt pretty stupid with 40% bonds in the portfolio. You might have raised your equity share, maybe to 75% or even 100%. Hey, what can possibly go wrong with pets.com and WebVan, right? Well, with the 75/25 portfolio you’d be down to $400k by now. With a 100% equity portfolio – favored by many FIRE bloggers, by the way – you’d be down to less than $160k. That portfolio with a current 25% withdrawal rate ($40k annually out of a $159,839 portfolio) will likely not survive for another 8 years.
Almost equally scary, even with a 60% equity portfolio but using a short-term fixed-income instrument (e.g., 3-months T-bills, money market, etc.) for the remaining 40%, you would have depleted the portfolio down to below $200k. That one’s also doomed even for a 30-year horizon. What would have been the rationale for the short-term fixed-income allocation? That was the allocation that worked really well in the 1970s when longer-duration bonds got hammered. So, in August 2000, to have a successful 30-year retirement, you had to ignore what worked best in the 60s, 70s and early 80s and instead foresee the long walk down in Treasury yields.
And needless to say, even with that perfect 60% stocks and 40% intermediate bond allocation, if you had retired in 2000 with a 50 or even 60-year horizon, then the 4% rule doesn’t look so hot for you today, and much less in 2009! In 2022, you’d have only $506k left, which implies that you now run an effective withdrawal rate of 40/506=7.9% with 28 to 38 years left, respectively. Time to cut your withdrawals significantly, probably down to 4.5% of today’s portfolio, which would translate into about $22,800. A 43% pay cut. Better hope that Social Security will make up the difference!
Why 2022 fundamentals look worse than 2000/2001
Another flaw is the idea that 2022 looks much more benign than the Dot-Com bubble peak. Let’s take a look at some of the financial and economic fundamentals at the two (month-end) market peaks: August 2000 vs. December 2021, see the table below.
It’s certainly true that stocks were more severely overvalued back in 2000. The Shiller CAPE stood at almost 43 compared to about 38 at the most recent peak. If you invert the CAPE ratios into an earnings yield (CAEY = Cyclically-Adjusted Earnings Yield = 1/CAPE), though, that’s not a meaningful difference. Today’s earnings yield of 2.61% is only 28bps higher than the 2.33% at the peak of the Dot-Com bubble. Of course, if we use the improved ERN-adjusted CAPE ratio, we notice a slightly larger gap of 0.83 percentage points, but that’s still not a significant enough difference to pop the champagne corks in early 2022, especially considering all the other factors that look a lot worse in 2022, such as:
- Rates: The 2022 interest rate picture, in contrast, is atrocious compared to 2000. The 60/40 portfolio post-2000 benefitted from the amazing diversification benefit of your bond portfolio. You started at almost 6% yields for the 10-year and then cashed in on the duration effect during the three recessions and bear markets in 2001, 2007/8, and even in 2020. In contrast, 2022 started with a 10-year yield of 1.52%. Not much room to walk that down, so even ex-ante this wasn’t a pretty picture. And ex-post, we know what happened: interest rates rose during 2022 and bonds didn’t offer any diversification benefit. In fact, year-to-date up to October 31, 2022, the S&P 500 and the 10-year Treasury bond index are both down by roughly 18% before inflation and 23% in real terms. And the 60/40 portfolio as well.
- Inflation: It’s not that inflation was particularly low in 2000. In fact, headline CPI and PCE were still a bit elevated. But core inflation was right where we want it: 2.48% for the CPI and even a little below 2% for the Core-PCE, which is the Federal Reserve’s preferred inflation measure. Compare that to 2021/2022. We faced not only high but accelerating inflation in early 2022. That’s bad for bonds, bad for stocks (at least in the short- to medium-term) and it erodes the (real) portfolio value in addition to the already poor nominal returns.
This combined rate+inflation picture meant that monetary policy is a great threat today, while in 2000 it was “Alan Greenspan to the rescue.” I remember the 2000 situation well because I started my job at the Federal Reserve Bank of Atlanta in September 2000. The Fed had just finished its rate hike cycle and declared victory over inflation. The central bank now had room to ease monetary policy and assist the economy during the slowdown. The opposite is going on right now. The Fed sat on its hands for too long, and instead of nipping inflation in the bud in the Spring of 2021, people pontificated about transitory inflation. Until inflation became pretty clearly self-fulfilling and much more permanent by early 2022.
So, if the 2000 market peak is not really the best comparison subject, is there a more comparable historical cohort? You bet, which brings me to the next point…
Let’s not forget: 2022 Looks much worse than the mid-1960s!
Another fly in the ointment of the 2022 cohort betting on things working out as “well” as in 2000: We have had certifiable failures of the 4% Rule in the mid-to-late-1960s when the Shiller CAPE was only in the low-to-mid-20s. For example, the CAPE was 24.04 in October 1965 and 22.78 in November 1968 and the 30-year safe withdrawal rate of a 60/40 portfolio dipped to 3.81% and 3.85% respectively. You don’t want to feel too safe when the CAPE is 30+. It took much less than that in the 1960s for the 4% Rule to fail! Today, with the Fed raising rates while the economy is on the ropes, looks so much more like the 70s than the late 1990s and the 2000 market peak.
But there’s some good news, too!
Before I send everyone on their way all sad and depressed, there are also a few positive signs. First, a bear market may not be as harmful to savers because the flipside of Sequence Risk is Dollar-Cost Averaging (DCA). Depending on how far you are from retirement and to what degree and how quickly equity valuations mean-revert, you may even benefit from a bear market by picking up stocks at a steep discount, right before another rally takes off. See my old post from 2019 “How can a drop in the stock market possibly be good for investors?“. So, for everyone not even retired yet – and that’s probably more than 50% of the FIRE community – just sit back and relax.
Second, as I wrote in my post earlier this year about retiring in a high-inflation environment, the level of inflation is less important than the direction. In other words, declining inflation rates, if they indeed come to pass over the next year or so, could be mostly positive for both the stock and bond market. Of course, some naysayers will object that inflation is stubbornly high and we’ll have to go through a Paul Volcker 2.0 experience. But being the eternal optimist, I cross my fingers that the Fed – while admittedly a bit late to the party – is certainly more responsive now than in the 1970s. If today’s Federal Reserve doesn’t drag out the inflation problem for a decade as in the Burns/Miller era (strictly speaking even starting under Martin), we should muddle through much faster and with much less pain and damage than back then.
Well, so much for today. After delivering mostly good news in the two October blog posts, i.e., today’s CAPE is not as astronomically high if we make the appropriate adjustments (10/5/2022) and with the more attractive valuations today’s retirees can also raise their withdrawal rates (10/12/2022), I just wanted to make sure I don’t sound too upbeat. I must defend my reputation as the retirement grinch.
Thanks for stopping by! Please share your thoughts and comments below!
Title Picture: The Scream by Edvard Munch