When you’ve finished this post you might also be interested in “The 4% rule can ruin your retirement. Don’t let it do your thinking for you!“
If your retirement plan involves a safe withdrawal rate, you gotta pay some homage to the guy who came up with the concept.
What might’ve been the most important moment in the history of the FI/ER movement got missed by almost everybody. It happened in October of 1994, when the Journal of Financial Planning ran a ten-page article entitled “Determining Withdrawal Rates Using Historical Data” by a California investment manager named William Bengen.
Bengen, a graduate of the Massachusetts Institute of Technology’s aeronautical engineering program, had run a family business for a while before moving out west, earning a master’s in financial planning, and starting a private practice. As he built up his client base he began developing a couple of ideas: “portfolio longevity” and the “maximum safe withdrawal rate”…a term he coined, and that’s still known as “The Bengen Rule.”
His findings proved to be an enormous boost to the FI/ER movement, and time hasn’t diminished their significance one bit. So I’d like to walk you through and comment on his points I consider most salient. But before we get started, do yourself a solid and print a copy of his article by clicking on this link.
Bengen’s theme was this:
“I have quantified portfolio performance in terms of ‘portfolio longevity’: how long the portfolio will last before all its investments have been exhausted by withdrawals. This is an intuitive approach that is easy to explain to my clients, whose primary goal is making it through retirement without exhausting their funds, and whose secondary goal is accumulating wealth for their heirs.” [Italics his.]
That first sentence is still the holy grail of the FI/ER crowd: amassing a self-sustaining portfolio by increasing sources of passive income and decreasing expenses. You can derive every principle of financial independence and early retirement from those few words.
After a brief executive summary Bengen set out a list of baseline market conditions drawn from industry-standard historical data. He calculated that across time the inflation-adjusted return an investor could expect from a 60/40 stock/intermediate-term Treasury bond split was 5.1% (assuming continual rebalancing.)
Bengen proposed, however, that a financial planner who recommended his clients withdraw 5% of their portfolio’s initial value could very well bankrupt them given the severe market shocks that came along every so often.
He listed the three shocks he considered “the most significant financial cataclysms of the last three quarters of [the twentieth] century”:
- The “Little Dipper” of 1929-31: -61% stock return, +10.5% intermediate-term bond return, and inflation of -15.8%.
- The “Big Dipper” of 1937-1941: -33.3% stock return, +16.7% intermediate-term bond return, and inflation of +10.5%.
- The “Big Bang” of 1973-1974: -37.2% stock return, +10.6% intermediate-term bond return, and inflation of +22.1%.
“As planners,” he wrote, “we know such events are likely to recur in the future.” So when advising clients, he felt it was crucial to take declines like these into account.
You’ll still see that approach in today’s financial calculators. While in his analysis Bengen didn’t use the more rigorous method of repeated random sampling–i.e, a Monte Carlo approach–the fundamental maxim remains: learn from history.
When choosing names for these cataclysms Bengen was obviously inspired by his astronomical studies. But he explained that the names also “reflect[ed] their relative impact on the value and purchasing power of investors’ portfolios.” The “Big Bang” was therefore the most severe: a investment pool eroded both in returns and purchasing power.
It was within the framework of these cataclysms that he outlined several hypothetical “portfolio scenarios.” Each described an investor who retired at a different date with a different asset allocation. All else was held equal.
Unfortunately, holding all else equal blurred the model a bit. Bengen expanded on his assumptions in an appendix to the article, and one that’s of particular importance to the FI/ER crowd was this:
“…the effect of taxes is neglected, as if all the retirement money were stored in a tax-deferred account. Analysis of a taxable account would be considerably more complex.”
Everything in tax-deferred accounts. Does this significantly diminish the usefulness of Bengen’s work for early retirees who live on pre-tax investment proceeds? I don’t think so, even though I’ve discussed the difficulty of managing taxable investments in a different post (“I Have a Big-Ass Tax Problem.”) We know that while these calculations are indeed complex, they’re well within the grasp of educated investors…especially given easy access to market data and the widespread use of spreadsheets.
A more serious problem with his assumptions is this: he never touches on the sapping effect of commissions, advisory fees, etc. The argument that these sources of friction can be ignored because they’d be constant across all portfolios is simply untrue–a 50/50 stock/bond allocation would have a significant difference in commissions than, say, a 60/40. And naturally investment advisors would charge different amounts and raise their prices over time, at variable rates.
At any rate, Bengen’s analysis was straightforward. He assumed a desired portfolio longevity of fifty years, which at that time was well-beyond the expected lifespan of a “traditional” retiree. He then presented bar charts showing–again, based on appropriate periods of historical data–how long a person who retired in a particular year and withdrew at a certain rate could expect his/her portfolio to last until complete drawdown.
Here’s Bengen’s chart for a withdrawal rate beginning at 4% in the first year:
So for example, at a 50/50 stocks/treasuries balance, retiring in 1926 was likely to result in a fifty-year portfolio longevity at a 4% withdrawal rate…whereas retiring in 1937 under the same conditions would only result in a portfolio longevity of thirty-nine years.
Bengen explained that, “[This figure shows] the effects of some financial events. However, these effects are comparably mild; no client enjoys less than about 35 years before his retirement money is used up.”
The chart’s damned near what you’ll get if you run your own scenario through FIREcalc: a graphic representation of your portfolio’s success rate.
Bengen found a 3% first-year withdrawal rate to be “as exciting as a crewcut,” with a 100% success rate for a fifty-year portfolio longevity. He also confirmed his proposition that a 5% first-year withdrawal rate could lead to a portfolio longevity of only twenty years.
Then he went on to describe the effect of each cataclysm on each withdrawal rate. But I won’t belabor these because his point was proven: portfolio longevity is directly affected when a client adjusts his/her withdrawal rate. That must’ve seemed an obvious conclusion even then, but presenting it and mathematically proving it using real-world data across varying historical timespans was still an important impetus to the FI/ER movement.
Bengen turned next to “Strategies and Applications,” and here we get into the meat of the article. As mentioned, he considered a 3% withdrawal rate to have a 100% success rate, but he qualified that by writing:
“This is also true for withdrawal rates as high as approximately 3.5%, but most clients would find such a low level of withdrawals unacceptable.”
Well…that might be true for traditional retirees, but it’s clearly false for the FI/ER crowd…especially when considered against his 100% non-taxable account restriction. I myself have run a withdrawal rate of between 2.75% and 3.75% (ish) pre-tax for most of the twelve years I’ve been retired. And this statement also dismisses those practicing leanFIRE.
Then Bengen repeats:
“Therefore, I counsel my clients to withdraw at no more than a four-percent rate during the early years of retirement, especially if they retire early (age 60 or younger.)” [Italics his.]
Defining early retirement as less than age 60 is intriguing. It implies a blind spot I’ll discuss more fully in my conclusion: prior to the internet era the FI/ER movement wasn’t anywhere near being in common knowledge…nor was it self-aware. Financial planners unaware of the 4% safe withdrawal rate could hardly be expected to properly advise prospective early-retirees, then, unless they spent a LOT of time on the math–which would’ve been difficult to do in the absence of PCs and spreadsheets.
To get around the difficulty of calculating numbers specific to the particular situation of each client, Bengen developed general outlook charts. For example, for a client with a $400,000 portfolio who wanted to withdraw 4% the first year and increase that rate to track inflation in each subsequent year, he would:
“…show the client…the chart for 6-percent withdrawals–and explain the risks of such an approach.”
“If the client expects to live another 30 years, I point out that the chart shows 31 scenario years when he would outlive his assets, and only 20 which would have been adequate for his purposes.”
A 40% chance of success. Again, a proto-FIREcalc, and another reason why Bengen’s work was seminal.
A final strategic point he made was this:
“…as we shall see later, a different asset allocation would improve this, but it would still be uncomfortable, in my opinion.”
So Bengen took up that topic in the next section of his article, “Initial Asset Allocation.” He combined forty charts of portfolio longevity arising from retirement date and withdrawal rate scenarios into one: a “Worst Case Portfolio Life.” Was this the first rudimentary Monte-Carlo analysis of retirement scenarios?
Here’s the chart. (Ibbotson, by the way, was a standard market data source in the early 1990s. Ibbotson Associates has since been acquired by Morningstar.)
Cool, isn’t it? Even given a worst-case scenario, a withdrawal rate of 3% is pretty much bombproof; at 4% and a 50/50 stock/bond allocation you still have a good shot at a portfolio longevity of 35+ years. Comforting to traditional-age retirees; perhaps less so for early retirees unless they manage their withdrawal rate down. Bengen did point out, however, that:
“From the perspective of the highest minimum portfolio longevity, that means you give up very little by increasing stocks from 50 percent to 75 percent of the portfolio.”
And then he made an observation I think is coolest of all, given my moderately high appetite for risk. He presented a chart of a 75/25 stocks/treasuries allocation which demonstrated that:
“Clearly, the heavier weighting in stocks…has produced some fairly significant improvements. Fully 47 scenario years [of 50] result in portfolio longevities of the maximum of 50 years, while only 40 scenarios result in portfolio longevities attained that pinnacle in the earlier chart.”
So he didn’t always consider a 50/50 stocks/treasuries mix optimal. After examining a few more cases of allocation/withdrawal rates, he concluded:
“I think it is appropriate to advice the client to accept a stock allocation as close to 75 percent as possible, and in no cases less than 50 percent. Stock allocations lower than 50 percent are counterproductive, in that they lower the amount of accumulated wealth as well as lowering the minimum portfolio longevity.” [Italics his.]
This held true even through the three cataclysms he’d first set out.
In the next section: “Asset Allocation and Withdrawals,” Bengen proposed something many might find controversial. Should retirees become more conservative in their allocations as they get older?
“My research,” he wrote, “indicates strongly that as long as the client’s goals remain the same, there is no need to change the initial asset allocation. It is likely to do more harm than good…”
To show this, Bengen divided retirees into three example classes; again, astronomically-named.
- The “Black Holes,” who retired exactly at the bottom of a cataclysm and were tempted to switch to a 100% bond mix.
- The “Stars,” who enjoyed…well…stellar returns, and
- The “Asteroids,” or those whose results were more or less what they expected.
There’s no need for me to summarize these examples; suffice it to say that Bengen was able to demonstrate historically across all three classes that maintaining a consistent 75/25 stocks/treasuries allocation in good times and bad led to the most consistently sustainable withdrawal rate of 4%, as well as leaving the most wealth behind for one’s heirs…no matter when one’s retirement date happened to fall.
At the very end of his article Bengen wrote something that plays into one of my long-held beliefs:
“The withdrawal dollar amount for the first year [of retirement] will be adjusted up or down for inflation every succeeding year. After the first year, the withdrawal rate is no longer used for computing the amount withdrawn; that will be computed instead from last year’s withdrawal, plus an inflation factor.”
I get behind this idea because as an early retiree for going on twelve years, I’ve found it’s wisest to model the future as if each day is the first day of my retirement.
I mean…you HAVE to look at it that way. Yesterday’s realities and assumptions aren’t today’s operating conditions; what matters is the likelihood today’s realities and assumptions will sustain your retirement in tomorrow’s operating conditions. If it looks like they won’t, you need a job…stat.
Today’s the first day of the rest of your life, yes?
As I get closer to wrapping up, this would be a good time to fess up that I’m a Bengen fanboy.
Reason being that in 1993, when I started my own planning, I had no idea this kind of thinking existed. And if I had, it would’ve not only given my model a boost but also would’ve helped me feel much more confident in the possibility I could actually achieve FI/ER.
Which goes to prove the point I made earlier (and in fact I make often): before the use of PCs and spreadsheets became widespread, and before the proliferation of the internet, it was right up against impossible for people who were interested in FI/ER to find each other and share information. But in a mere twenty-five years knowledge of the 4% rule has gone from being obscure to widespread–hence why I call the FI/ER movement a “movement” in the first place.
And that’s why I’m a Bengen fanboy. Somebody had to be the first to bring the 4% rule to light, and while in the introduction to this paper he credits other financial planners for laying in some of the conceptual framework, he was the one who developed and presented the actual mathematical method that seekers of FI/ER could easily apply to their own situations. I could make the case that the majority of subsequent retirement modeling has been influenced by his work. Certainly the Trinity Study was predicated on it.
But naturally these kinds of safe withdrawal rate methods have come under considerable criticism. The most common I run across is this: retirement spending shouldn’t depend on a blanket withdrawal rate, but rather on a percentage of pre-retirement spending that varies by individual. I’ve seen recommendations as high as 85%.
And I think that’s crazy. Recommendations that high completely ignore the “push down on your expenses” side of FI/ER; as if retirees have no self-discipline. I’m committing the logical error of attacking the argument by ridiculing it, of course, but I still think Bengen’s work convincingly demonstrates that given external marketplace realities, investment sustainability should be the main goal…especially to an early retiree. That suggests managing one’s expenses down as much as possible.
Another criticism I hear is that while a 4% withdrawal rate is “smooth,” the underlying markets are “choppy,” meaning retirees build up surpluses during high-return periods and run deficits during low-return periods. It’d be much more appropriate, these critics argue, for one’s withdrawal rate to vary with market returns.
But that doesn’t jive with the time-honored method of how commodity companies manage the price risk of their income streams. Say you’re a wholesale power company running coal generation. Since you stockpile weeks’ worth of coal but sell into a spot market for electricity, you’re engaging in a fixed-for-floating swap, or at least in the short term. Fixed expenses; variable revenue. So you hire futures traders and sell into the forward markets to lock in longer-term prices and smooth out your income stream. Investors like this because certainty’s always a good thing in the stock market.
Consumers and especially retirees are likewise caught in a fixed-for-floating swap. Their income remains constant but their bills fluctuate. That’s uncomfortable, and something that has to be carefully planned for…especially in times like Bengen’s Big Bang, when abysmal returns team up with high inflation to smack retirees all over the playing field.
Which is an argument for buying TIPS and annuities, maybe. As of this writing TIPS are paying as much as 3.875% depending on how far you go out. I’m a long way from 67, but I’m becoming more and more convinced that Social Security will be around for me…and if I were to buy a million bucks in TIPS at 3.875% and combine that with a couple of grand in combined monthly Social Security payments for my wife and me, I think we’d be able to maintain our current more-than-comfortable lifestyle, no problem. Antithetical, surely, to Bengen’s conclusions with respect to passing wealth on to one’s heirs, but an interesting question nonetheless, and certainly within the boundaries of a 4% safe withdrawal rate.
Well…I’m going to leave it there. I think I’ve shown that not only was Bengen’s work seminal to the FI/ER movement, and not only does it continue to be of high importance, but it also brings up all sorts of tangential questions that themselves are worthy of thought and study.
And that said, another tangential question is this: I had a six-pack of Stella Artois around here somewhere…where the hell did it get to? I’ve been working on this article for nine hours straight and I need to wet my throat. And Wikipedia tells me “Stella Artois” means “The Christmas Star” in French…surely a name that’d please a financial star-gazer like William Bengen.
By the way, these kinds of historical posts are common at my blog. Please click here to follow me on Twitter, and sign up for email updates using the widget in the sidebar. Thanks!