The First Question You Should Ask About In-Retirement Withdrawal Rates
Striking the right balance between what's sustainable and what's livable is job one.
Is 3% the new 4%?
If you follow retirement-planning matters, you've probably read your fair share of articles about that topic--and about the right withdrawal percentage generally. Several recent research pieces, including one from my colleague Amy Arnott, have asserted that 4% is probably too high, given the confluence of today's low bond yields and not-inexpensive equity valuations.
That's an important point. But I'd argue that homing in on a specific percentage skips a key step in the process of crafting a durable withdrawal system. A better starting point when thinking about portfolio withdrawals is to determine your own tolerance for variation in your in-retirement cash flows.
Are you seeking a withdrawal system that's much like the paycheck you earned while you were working--a fixed real-dollar withdrawal each year? Or are you OK with annual fluctuations in your annual withdrawal amounts, in line with your changing portfolio balance, if it helps improve the odds that your portfolio will last throughout a 25- to 30-year retirement?
There's no single right answer; there are trade-offs with both approaches. But figuring out how you'll balance sustainability with livability is a key first step when thinking about withdrawal rates.
As retirement researcher Wade Pfau framed it in a conversation on The Long View podcast this past year, it can be helpful to think of withdrawal systems on a spectrum.
On one end is a static real-dollar withdrawal system; that's the framework that underpins the 4% guideline. Such a system assumes that the retiree withdraws a percentage, such as 3% or 4%, in year 1 of retirement, then inflation-adjusts that dollar amount thereafter. It delivers a stable inflation-adjusted cash flow from the portfolio. To use a simple example, a retiree using a 3% initial withdrawal rate on a $1 million portfolio could take out $30,000 in year 1 of retirement, $30,900 in year 2 (assuming 3% inflation), and so on.
Bill Bengen used that setup for his seminal research on withdrawal rates in the 1990s, looking back on market history to determine a starting withdrawal percentage that would have enabled a balanced portfolio to sustain itself over a 30-year period. He hit on 4% as having been sustainable over even the worst 30-year period in market history, though more recently researchers have suggested that a lower starting withdrawal is advisable today.
On the other end of the spectrum would be a fixed-percentage system where withdrawals depend completely on the portfolio's balance. If the ongoing withdrawal rate were 4%, for example, the retiree with a $1 million portfolio would pull $40,000 in year 1. Assuming a 30% market drop in year 2, she'd be living on $26,880 in year 2. If her portfolio went up by 30% in year 2 rather than down, her year 2 withdrawal would jump up near $50,000.
By now you can probably see the major positives and negatives with both of these systems. Taking fixed real-dollar withdrawals--Bengen's 4% approach--has the advantage of simulating a paycheck, which many retired people take comfort in. But retirees' spending may not always move in a straight line; retirees have high spending years (new roofs, big trips) and low ones (hello 2020!). More important, this type of system isn't at all sensitive to the underlying portfolio's value. A retiree may start by taking out $40,000 from her $1 million portfolio (4%). But if her portfolio drops by 30% in her first year of retirement, her year 2 withdrawal of an additional $40,000 spikes up close to 6% of her balance. If her portfolio continues to decline and she doesn't adjust her withdrawal rate, she runs the risk of prematurely depleting her portfolio. That risk is exacerbated if her portfolio is very conservative (and hence its return potential is poor), the bad returns arrive early on in her retirement (sequence risk), or if she has a longer time horizon than the standard 25- to 30-year period.
Meanwhile, the other extreme--taking a fixed percentage year after year--nicely addresses the sustainability problem. By taking less in year 2, when her portfolio is down, our hypothetical retiree leaves more to recover when the market eventually does. Her withdrawals are only larger when her portfolio can support them. But her cash flows are buffeted around from year to year more than would likely be ideal. In good years, she has more to spend, but in lean years, she may not have enough to cover living expenses or outlays that constitute her quality of life, such as travel and dining out. That variability can make planning difficult and add stress.
The good news is that retirees don't have to choose between these two extremes. Numerous flexible withdrawal systems try to balance the two major sets of interests, helping retirees maintain a somewhat steady portfolio paycheck while also tethering their withdrawals to what's going on in their portfolios.
The most basic such system would be based on required minimum distributions: Withdrawals would depend on the prior year's balance, but they would also increase a bit annually to keep pace with the retiree's reduced life expectancy.
Pfau notes that he also likes another method that Bengen developed, which uses fixed annual percentage withdrawals as a starting point, but puts a "ceiling" and "floor" underneath the amounts. That way the retiree isn't forced to radically rein in spending after bad market years, and she can't go too crazy with overspending after a strong stretch like 2019-20. Financial planner Jonathan Guyton has developed a strategy along these lines called the "guardrails" system.
Customization Is Key
As you think through whether fixed real-dollar withdrawals or a more flexible system is the way to decumulate during retirement, you may know right away which setup is most appealing to you. But here are some additional questions to help you think through where your withdrawal system falls on the spectrum.
Do you have a high risk of running short of retirement assets?
Yes: A more flexible system--specifically, being willing to rein in spending if a weak market occurs early in retirement--will help improve your plan's odds of success.
No: You can put more weight on the peace of mind that comes along with having a stable stream of cash flow in retirement.
Do nonportfolio assets (Social Security, pension) cover a lot of your in-retirement living expenses?
Yes: A more flexible withdrawal system will be easier to live with. (Fluctuations in withdrawals won't affect your basic living expenses.)
No: A static dollar withdrawal system (or something close to it) won't jeopardize your ability to pay basic expenses.
Are you willing to be flexible in terms of your discretionary spending in retirement?
Yes: A more flexible withdrawal system could be workable if you're willing to change your spending plans based on how your portfolio is behaving.
No: A static withdrawal system will allow for greater predictability if you like to plan ahead.
Did you experience a lot of variability in your paycheck during your working years? (For example, much of your salary owed to lumpy income sources like bonuses and commissions.)
Yes: A more flexible system could be workable; you're used to this!
No. A static dollar withdrawal system will feel more comfortable and allow for planning.