January 13, 2021
Happy New Year, everyone! And welcome to a new installment of the Safe Withdrawal Rate Series. Today I like to write about the One More Year Syndrome (OMYS) – the fear of retirement and the decision to just work another year. What I find intriguing about OMYS is that procrastination normally works the other way around. You opt for the fun and easy stuff and promise yourself to do the hard work tomorrow. Only to repeat that charade again tomorrow and postpone the unpleasant tasks to the day after tomorrow. And so on.
But why procrastinate a fun-filled early retirement and keep working? Physician on FIRE and Fritz at The Retirement Manifesto have written about their rationales. The number one reason is that you grow your nest egg and put your retirement finances on a better footing. That was certainly my main rationale, too. I could have retired comfortably in 2017, probably even in 2016 but I delayed that decision until 2018.
So, qualitatively it’s obvious. But can we quantify by how much the OMYS improves your retirement security? Is it worth the additional year in the workforce? How can we incorporate OMYS in the Big ERN Google Safe Withdrawal Simulation Sheet? Is it possible that OMYS will boost your retirement health so substantially that it’s not as irrational as it’s sometimes made? Let’s take a look…
Let’s start with the following parameters, to be fed into the Google Sheet (See SWR Series Part 28 for more details):
- A 30-year retirement horizon (stay tuned for longer horizons later).
- A 75/25 portfolio, i.e., 75% in a U.S. S&P 500 index fund (e.g. iShares IVV), 25% in a U.S. Intermediate Treasury ETF (e.g. iShares IEF).
- No additional cash flows along the way. (but there will be a version with Social Security benefits later)
- A $2,000,000 initial portfolio value. The exact dollar value is irrelevant because all numbers will scale linearly. I just like to use a round number for illustration.
- A final portfolio value target of 25% (i.e., $500,000) after 30 years.
The reason I like this scenario as the baseline is that it easily applies to both early retirees and traditional retirees. This could be a traditional retiree with a 30-year horizon, where all the supplemental flows from Social Security and pensions are treated separately and you simply add that to your retirement budget. And the retiree has a bequest target of leaving a quarter of the initial nest egg – adjusted for inflation – to their heirs. Or one can think of this as an early retiree at age 35 or 40 trying to bridge the 30 years until Social Security and company pensions start that are almost sufficient to cover the budget then. Hence, the 25% final value target to supplement that late-stage retirement.
Let’s see how the safe withdrawal rates and safe consumption amounts look like in this example. I always like to monitor the failsafe consumption amounts, both overall but also by decade to see in what environment the really bad outcomes materialized. In the tab “Cash Flow Assist” I display a table “Failsafe by Decade” to do just that, see the copy of that table below. Notice how in this example, the 4% Rule is clearly not safe. Both in 1929 and the 1960s you had multiple occasions where this rule would have failed you. The safe withdrawal rate was closer to 3.6%, i.e., around $72k for a $2m initial nest egg.
So far, so good. Let’s see how much of an impact you’ll get from delaying retirement by one year…
Modeling the “One More Year”
I assume that in the “One More Year” case, our retiree can contribute $5,000 per month to the nest egg. So, the boost from the OMY comes not just from the one year reduced retirement horizon but also the additional contributions while working for another year. This raises the big question:
How do we implement this “one more year” scenario in the Google Sheet?
I can think of two ways. First, one could make “some assumptions” about the portfolio returns over the first 12 months and then tag on a 29-year retirement horizon simulation with that projected portfolio value. I don’t think that’s the most appropriate and elegant way, though. What I’ll do in this post here is to use the tools in the Google Sheet (See Part 28 for the details) and use the historical asset returns not just during retirement but even during the OMYS period. So, there’s no need to “make up” asset returns during the OMYS phase.
Now, how do we implement the two twists, 1) additional contributions and 2) the lack of withdrawals during months 1-12? Very easy. Let’s go to the tab “Cash Flow Assist” and…
- Enter the $5,000 as a positive number in the supplemental cash flow column.
- In column P we can apply a “scaling” to the withdrawals and we simply set the values to zero during months 1-12. (For the math geeks, check out Part 8 of the series, where I derived the mathematical formulas used in the SWR simulations, including the ones that utilize the scaling of withdrawals. The scaling feature allows you to model all sorts of neat non-standard, non-Trinity-style withdrawal patterns. For example, the OMYS or consumption profiles that grow x% faster than inflation, or many other other bells and whistles)
- And also, I set the initial portfolio value to $2,000,000. Notice that this has to be the initial value before the OMY because the portfolio value upon retirement, 12 months into the simulation, is now a random number; it relies on uncertain historical returns.
How much of an impact does the OMYS have on our failsafe withdrawal amount? Let’s do the OMYS adjustments in two steps. 1) delay retirement but hold off on the additional contributions and 2) in addition to the delayed retirement also apply the additional contributions.
Here are the results, see the table below. Delaying retirement without any new contributions already gives you an increase in the failsafe amount by more than 4%. The additional contributions bring you to just under +7.8%. Pretty impressive!
We can also simulate the longer horizon of 50 years, though (for now) without any additional cash flows (Social Security, pensions) and with a zero final value target. See the results below. Despite the zero final value target, you’ll get a significantly lower failsafe consumption amount ($67,874 vs. $71,683) in that baseline. Longer horizons simply require a more cautious withdrawal rate. I have been preaching this since 2016. See Part 2 of the series! Distilled in the title image of that post:
But the percentage impact of the OMYS is roughly the same: A boost of about 4.2% from delaying retirement and just under 7.9% for both effects – delay retirement and additional contributions.
50-year horizon plus Social Security/Pensions starting in year 30
And finally, let’s also add the scenario with the 50-year horizon but with a $3,000 a month Social Security and/or pension benefit starting in year 31. Quite intriguingly, this scenario gives you similar failsafe withdrawal amounts as the 30-horizon plus 25% final value target ($72,031 vs. $71,683 in the baseline without the OMYS). Which was exactly my point at the beginning of this post: a 30-year horizon with a substantial final value target is not that different from a 50+-year horizon with some substantial Social Security and pension benefits later in retirement.
And the OMYS gives you a similar boost in the failsafe consumption amounts: +4.18% (delay retirement) and 7.53% (delay retirement + $5,000 monthly contributions). Notice that even with the OMYS you’ll still remain under $80,000 annual safe withdrawals, as implied by the 4% Rule. If we push the OMYS to a Two-more-years-syndrome, we’ll certainly get above the $80k mark with a 15+% boost in the failsafe consumption amount. (actually, the cross-over point was at 18 months to make the 4% Rule 100% safe).
Lessons learned so far:
- The 4% Rule was not safe in historical simulations, not even with a one-year delay in retirement and additional contributions.
- The impact of OMYS on the safe withdrawal amounts is surprisingly uniform. Whether it’s a 30-year horizon or 50-year horizon, whether it’s with or without additional cash flows, you boost the failsafe withdrawal amount by about 7.5-8%.
- The part of the extra failsafe consumption that’s coming solely from delaying retirement is about 4%. That’s a big impact because in the case of the 50-year horizon you reduced your withdrawal duration by only 1 year, i.e., only 2% of 50 years.
OMYS with a 4% fixed withdrawal rate
Another interesting exercise: instead of raising the sustainable withdrawals we could also fix the withdrawal amount, say, at $80,000 a year or 4% of the initial net worth, and then simulate the portfolio value over time with and without “One More Year” to see how the OMYS reduces your failure rates of the 4% Rule. That’s what I do in the table below. I use the same scenarios as above, but now the results for each scenario are in the rows of the table.
As always, I want to caution against using the naïve unconditional failure probabilities. See the table below. For example, if we look at the 30-year baseline, the unconditional failure probability is 6%. But conditional on the Shiller CAPE below 20, the 4% Rule failed only 2% of the time, while at a CAPE>20 you fail 18.8% of the time. Most of the failures of the 4% Rule are clustered during the times of very expensive equity valuations. Some people can live with a 6% failure probability (I couldn’t). But with today’s CAPE north of 30 and an implied historical failure probability of 18.8%, it seems like a no-brainer to reduce the withdrawal rate to less than 4%.
OMYS significantly reduces that failure probability to 0% in low-CAPE environments and 4.3% when the CAPE is above 20. Also, notice how the conditional probabilities are very similar to the 50-year baseline with Social Security, which is why I really like the 30-year baseline with a 25% final value target because it has a very similar risk profile to the longer horizon with substantial Social Security at the end. Also notice that in the 50-year case without Social Security, the OMYS helps reduce the CAPE>20 conditional failure probability but it only goes from 31.8% to 18.3%. Still quite scary!
A similar picture, qualitatively, emerges when you condition on the “relative” historical valuation of the stock market, i.e., where we are relative to the all-time-high (up to that point). Failure probabilities are in the double-digits if the S&P 500 is at or within 10% of the peak in all the baseline scenarios (without OMYS). But essentially zero when you’re more than 10% under the equity all-time-high. With the OMYS assumptions, you cut the failure probabilities by more than half and with two more years, all probabilities go down to zero.
The OMYS is often portrayed as an irrational fear of retirement: Procrastination of work-addicted corporate slaves. But there is some reason for this insanity: a relatively brief delay of retirement will add a lot of additional safety margin to your retirement. You could either significantly increase your retirement budget or, for a fixed budget, you’d seriously reduce the risk of running out of money. Call it pre-retirement flexibility! It’s probably easier to pull off than the often-touted post-retirement flexibility. As I showed in Part 23, cutting your retirement budget and/or going back to work in response to a bad Sequence Risk event can last much longer than just the duration of the Bear Market. Hence, my decision to delay retirement until 2018 to grow the portfolio to over 30x while working in a cushy corporate job. I was also reminded of the case study for Mr. Stop Ironing Shirts, where just 9 months of extra employment put him on a much safer path through retirement, thanks to some deferred compensation vesting at that later date. He’s now retired, check out his blog, and seems to be happy with the decision to work just a little bit longer.
But just to be sure: don’t overdo the OMYS! I always like to rationalize that with the beautiful Dietrich Bonhoeffer quote:
“Time is the most precious gift in our possession, for it is the most irrevocable”
In that 2017/18 case study series I certainly urged several case study subjects to retire ASAP because the extra contributions didn’t add much to an already very safe retirement picture. Personally, we count our blessings every day knowing that we retired when we did. In 2018, we were able to travel for 7 months and then for another 4 months in 2019. With our daughter in school now and pandemic travel restrictions, we will never have that liberty again anytime soon. Be sure to carefully weigh the benefit of more retirement security against the lost time!
Thanks for stopping by today! Please leave your comments and suggestions below! Also, make sure you check out the other parts of the series, see here for a guide to the different parts so far!
Title picture credit: Pixabay.com