by Michael
It’s about time again for another post in my Ultimate Retirement Guide Series. As a refresher, the goal of this series will be to provide an in-depth answer to this question: “what do I need to do to retire?” As well as any questions that could stem from that initial question. The goal is to have this cover all of the information I’ve looked at since I started really becoming interested in retirement saving, both early or traditional. I hope to mostly focus on items that have been relevant to me in my own experience and where I am in my journey at the time of writing. Throughout, I will try to present as unbiased a position as possible, but I’m sure that I will not be perfect.
So far, we’ve laid out the following five large steps needed to retire:
- Determine when to retire.
- Determine the amount needed to retire.
- Save that amount.
- Prepare to transition from working to retirement.
- Retire.
Currently we’ve been working through step 2 by breaking it into the following subsections.
- The 4% rule of thumb
- Other withdrawal rates
- Other withdrawal strategies
- Length of retirement (and sequence of return risk)
- Inflation
- Asset Allocation
- Different types of investment strategies (rentals, stocks, ect.)
- Other expenses
- Visualizing retirement
Last time we discussed different withdrawal rates and strategies, and my opinions on each. One topic that keeps coming up, is the length of the retirement and how that can affect some of these topics. This time we’ll talk about the length of retirement and the biggest risk a retiree faces; sequence of returns risk.
My usual disclaimers apply to both things we’ll talk about here. I’m not an actuary nor am I an investment professional. Everything on this website is opinion and should not be taken as advice. Seek professional advice for all changes to your portfolio or plans.
Length of Retirement
My biggest criticism of the 4% rule as gospel for retirement is that it does not account for the length of retirement, let’s talk a little bit about some approaches to determine that length.
There are exactly three things that are important to determining the length of retirement: age at retirement, estimated age of death, and a contingency. Age at retirement is almost entirely defined by the individual and their plans, in my case I’m planning for 35.
Age of death is a little bit harder. Using the Social Security Administration’s Life Expectancy Calculator, it looks like I’m likely to survive to 81 today (source: for a 23 year old, male). If I make it to my 60s, that extends to my mid-late 80s at 87. Let’s plan for that number.
How likely is it that I may overshoot that estimate and live longer? I don’t know but if I don’t at least consider it I may plan for too short of a retirement and be underfunded at the end of my life. The above number for my age of death is based on my cohort, less so on my individual lifestyle and family history. I know that my family tends to live a little longer than average and I individually try to take care of myself, so let’s round that number off to 90. I should also add something in for other items I may not have already considered or that are harder to quantify, like medical advancements, so let’s add another cushion of 5 years to get a final age of 95.
To review real quickly for my specific numbers: I plan to retire at 35, I plan to live to about 87, and I have a contingency of about 8 years. My overall length of retirement then is about 60 years (starts at age 35 and ends at age 95). Some of the assumptions another individual uses could change this number a bit, say considering family history or lifestyle factors, maybe just a lower contingency for unknown factors. You could make a fairly convincing case to me that I may make it to my late 90s.
Anyone who has been reading the other stuff I’ve been writing about on how much someone should be saving should know that the 4% rule was designed to plan for a 30 year retirement and has exceptional results for that length. In my case at a 60 year length it has a lower chance of success and I was not comfortable taking that kind of risk, so I’ve adjusted my withdrawal rate accordingly.
While for some thinking about their own death may be uncomfortable, I believe that retirement planning can’t be done without at least a ballpark idea of when you’ll die. It doesn’t have to be perfect but a good enough estimate with a bit of a contingency in case you exceed your expectations will give you a good enough number to work with.
Sequence of Return Risk
Speaking of time in retirement, it’s time we talk about what is often considered the greatest threat to a retirement, traditional or early, sequence of return risk.
Usually, the sequence of return risk is portrayed as the following: early in retirement (usually the first 5 years) there is a downturn that lasts for a number of years; this causes the retiree to have to sell more assets than initially planned. Those additional assets can no longer compound and may force the retiree to have to adjust their withdrawal rate downwards to compensate.
While saving for retirement the order in which you make your contributions doesn’t really matter compared to your average rate of return over that period (assuming near constant contributions). Yes, being able to buy more shares at a younger age is great but it has a smaller effect in the grand scheme of things compared to the amount saved.
However, in retirement the order matters a lot more. If you have 100 shares and sell 6 of them to cover your retirement when you initially needed to sell 4 per your plan the compounding from those remaining 2 is gone forever, add enough of those up and your retirement fund may disappear.
Another problem with the Sequence of Risk Return is that it never truly goes away. A large enough or long enough downturn could cause failure no matter how many years into retirement you are (other than probably the extreme tail ends). Would your retirement plans last 60 years with a repeat of the Great Depression 10 years into those 60?
This is not to say we should throw our hands up and say “I’ll just take that risk that if a long enough or bad enough downturn occurs I’m just finished.” Good personal finance is not ignoring risk or just hoping it won’t be your problem, but instead weighing them. For instance, complete global economic collapse in my lifetime is possible. I believe it is extremely unlikely, plus the consequences for me specifically of being wrong mean little if that were to happen (they’re even asymmetric if it doesn’t happen). A similar thing can be done with less extreme events, is a repeat of the Great Depression possible in my lifetime? Yes. Is it likely? I think less so. We’ve learned a lot about how economic markets work since the 1920s and 30s, but we have not learned everything nor do I think we’d ever be able to, so I plan for Great Depression-like events.
On the other hand, I think certain upward markets are less likely to be repeated in my lifetime. Another individual, of course, may come away with slightly different likelihoods of similar events, you then take these likelihoods and adjust your plans accordingly. For instance my plan should be able to withstand even a robust downturn at any point over that 60 year period except for me happening to retire at the absolutely most extreme examples such as on or a few days before Black Thursday.
Another risk to consider is that if this event hits later in retirement compared to earlier it may be harder to find a job. I can say I took a sabbatical if I need to go back to work in the first few years, that may be a lot harder after 20 or more years.
One last note of the sequence risk is that it actually can benefit you in your accumulation years. If the market got stuck in a downturn for the first five years of accumulation (or at my peak 5 earning years) that could allow me to buy a larger number of shares compared to my initial plan and have a longer time for those extra shares to compound.
Now, you may be wondering how do I make these plans? I stress test them. There are plenty of retirement calculators you can plug your plan into and see if it will work (I like Fidelity’s a lot although it doesn’t allow for trying different withdrawal rates), this is great for early planning. There are also some calculators out there that crunch the numbers solely off of the historical data, I think these are better if you want to see if your portfolio would have survived say the last 100 years or compare retiring at the peak before the Great Depression to retiring 5 years earlier. Usually I prefer the historic data to the Monte Carlo data since the latter is a lot easier to manipulate.
To summarize, two of the many things that need to be considered while figuring out how much to save for retirement are how long that retirement will last and how to survive a prolonged downturn during said retirement. For one you need to consider your planned retirement age, age at death, and a small contingency. To survive downturns you need to plan for some and adjust your plan based on how likely you think these events are.
