FREE NEWSLETTER
Go to main Forum page »
This floated across my screen a couple of days ago.
Bill Bengen introduced the 4% rule in 1994. It suggested that a one may safely withdraw 4% from a portfolio in the first year of retirement and it would be likely that portfolio would last for 30 years.
Bengen is publishing a new book A Richer Retirement: Supercharging the 4% Rule to Spend More and Enjoy More
In it he updates the rule but also provides a 55/45 model portfolio. This is my understanding:
Bengen states that slightly overweighting small-cap and microcap stocks could “increase [the safe withdrawal rate] maybe a quarter of a percent, which is worth it.” He acknowledges that small- and microcap-stocks are very volatile classes and should be used in moderation. My portfolio includes small caps, and I’ve always wondered what an appropriate allocation would be. With Bengen’s allocation I have more info on this.
The new initial safe withdrawal rate has been increased to 4.7%. In practice the SWR could be increased by the rate of inflation each year. In the article I read he also suggested that a reasonable starting point for withdrawals today would be between 5.25% and 5.5%. This is lower than the historical average withdrawal rate of nearly 7%. Bengen acknowledged that “you’re giving up a lot, but those are the circumstances we face.”
He was quoted “The 4.7% is still a ‘once-in-a-century worst-case scenario’ number,” he said, and retirees “should probably be more optimistic than that.” He considers current inflation to be moderate and stock market valuations are “very high.”
Bengen uses the Shiller CAPE ratio (Cyclically Adjusted Price-to-Earnings ratio) to measure valuations. He noted that while the historical average is about 17, today’s CAPE is nearly double that, in the mid-30s. “I think if you get much above the low 20s, the risk starts to build,” Bengen said, warning that historically, high CAPE ratios have not lasted long and typically preceded sharp declines.
Here’s a brief synopsis of the book over at Amazon:
I know this question isn’t about the 4% article but, I do have a retirement income question.
We’ve always invested our retirement monies into Index ETF’s.
We have entered retirement with enough income from two SS accounts and three pensions to cover all of our expenses.
If I pass away first my wife will lose my pension and might need some form of extra income. She isn’t very interested in finances. Would it be easier for her to make a monthly sale if funds to provide it or for me to consider putting about seventy five percent of our retirement accounts into dividend ETF’s,Reit’s,etc reinvesting the dividends until a time when she could start collecting the dividends if they are needed.
I know most of us HD readers are indexers but I am asking it is curiosity
There’s a bunch of ways to create income from index ETFs. You’ll need to figure out what makes the most sense for you and your wife. The best thing would be for her to get up to speed on your investments so she can make this decision if you die first. Sounds like you’ve done great so far. Good luck!
Thank you for commenting, Ben.
I can’t imagine being comfortable spending 5% of my portfolio/year in the first few years. What if you retire into a down market? That money is gone. I plan on 2 – 3%. I have about 65% in stocks, the rest in near cash investments – very, very little in bonds.
I’ve seen dozens of discussions of Bengen’s work and one of the most common themes I see is folks translating it into an actual personal retirement withdrawal strategy. As retirement expert Wade Pfau says, it’s research, not a retirement income strategy. The right way to use information like this is simply as a reference point in designing a household’s retirement strategy which is typically based on more than one income source including Social Security. Lots of folks don’t use a straight safe withdrawal rate (SWR) approach and do fine.
I agree, it is research. I have used a variety of tools to reach a conclusion. These include a Monte Carlo simulator for back testing multiple portfolios, a “Lifetime Planner” which includes long term budgets as well as projected returns and also Benz’s Morningstar approaches (guardrails, straight line, inflation adjusted, etc.). My goal was to construct a number of scenarios and approaches that gave me likely outcomes and overall positive results. That included spending at a rate that allowed me to do everything I wanted while in retirement. One has to start somewhere.
Several of the comments have pointed out the drawbacks of using static models (such as the SWR) to plan for retirement spending. The Guyton “guardrails” approach allows for variable spending depending on market returns but has been described as cumbersome to work with. Economist, Ben Mathew has introduced a free calculator (it does require you to register an account) that uses a dynamic strategy, Total Portfolio Asset Withdrawal.
TPAW Planner
With fairly basic data input it runs 500 Monte Carlo simulations using your chosen allocation profile to give a monthly spending amount from the first month of retirement until the end date of your choosing. It provides a median figure as well as 95% confidence limits depending on market variability. The utility of the methodology is that it provides a broad picture of possible consumption scenarios in retirement and allows you to make on-the-fly adjustments in savings rate, retirement date, extra spending, legacy goals, and a monthly spending ceiling (or floor).
It is not as detailed as some financial planning software. For instance, it doesn’t try to tackle the problem of where the withdrawals are coming from and the tax consequences. It doesn’t track your particular portfolio, rather, it takes your asset allocation as a percentage of domestic stocks and bonds and reflects the performance of that mix. I have found the tool to be useful to get a general idea of expected retirement consumption, and based on the comments here, it may be a useful tool for some people to explore.
As I read through all the comments, I am struck, again, by the diversity of opinions regarding portfolio construction.
Everyone, including Bill Bengen, has their theories and reasons for their portfolio make up, and they are all interesting. But in reality, all the guess work and research and planning is nothing but market timing in disguise.
VTI and VXUS and BND and BNDX equals everything, domestic & international, except cash.
The only question is what percent of each of these four ETFs to you use to construct your portfolio?
OR…if you have fixed income covered by annuities, just VTI and VXUS, and some cash, which is what I hold.
Simple, efficient, time saving, and very easy to manage.
I have read all the comments. I would say it’s more important to save a substantial portion of your income, and invest in something. What particular theory of investing you follow is up to you, and it should be whatever your are most comfortable with. You can be a value investor, a growth investor, a dividend investor, an index investor – it doesn’t matter, you will make money. If you don’t make as much as you could have with some other method, you still have plenty of money.
Tying to “out perform” market returns is a fool’s errand for me and brings in too many emotions. Simplicity: Total US Market Index, Total International Market Index, Intermediate Treasury Bond Index, Short Term Bond Index and Cash, 3.5% withdrawal rate, fits my personality.
When I set up our retirement income plan in late 2023, ahead of my January 2024 retirement, I decided to use 4% for the first 2-3 years.
Our SS benefits and Annuity Income combined, equaled 184% of our retirement expenses. Adding the 4% from our portfolio made our Total Income 221% of our retirement income.
Could we take 5.5% or 6% SWR? Sure…but we don’t need it.
With a couple of trips planned in 2026, we may increase it for a year in 2026, but 4% seems to be more than adequate for our needs and desired lifestyle.
I don’t really like that portfolio. That’s a lot of microcaps and not a lot of Int’l exposure. Just my 2 cents, I prefer getting my diversification outside of the US, in case there are currency risk and other unforeseen circumstances. (who saw Japan becoming the size economy it did).
My regret of my own portfolio is that I have too many holdings and I’m working on whittling it down.
What I do like about Bengen’s update is the model portfolio to support his withdrawal thesis. The brief I read stated that the overweighting of small- and microcaps may add 0.25% per year to the withdrawal rate, but with higher volatility. I’ll probably read the book to get a better understanding of how he came to that conclusion.
Is owning more small- and micro- caps worth achieving that possible gain? Looking at the most recent 10 and 20 years performance, Large- cap and Small-caps were ranked 1 and 2 for investment returns in a group of nine key indices (per Callan). Developed Ex-US equity came in 4th and 3rd over the same periods. Callan indicates that over the 20 year period U.S.Large-caps returned 237.59%, U.S.Small-caps 232.67%, and Developed Ex-US 171.35%. Emergin Market Equity lagged. I’m aware that there has been renewed interest in Developed Ex-US equities. However, this has returned only 74.12% for the 10 years 2015 to 2024.
I have 30 years of financial data for my retirement plan; that indicates what has worked for me. I last made a significant change in 2020-2021 when I re-allocated to what I considered to be an appropriate “in retirement” portfolio. My 30 years of history included some significant events such as the Dot-Com bust, the 2008 financial meltdown, Covid, etc. It is difficult for me to consider making additional changes but to add simplification. However, my spouse’s horizon might be 25 years distant.
If I compare to Bengen’s allocation, I and my spouse’s combined portfolios have more large caps, in part because we like the dividends and stability. Our small- and micro-caps are lower but our foreign stocks are slightly higher. It would be relatively easy to increase my microcap exposure via Vanguard’s VTWO Russell 2000 ETF, or Wisdom Tree’s DES WisdomTree US SmallCap Dividend ETF. However, VTWO has done better.
Anyone who reads this please understand, I am not making an argument against small caps, I’m making the argument that there are divergent views on the future of small caps (and whether we can expect to see a real total return advantage that justifies the added risk and std deviation experience associated with small and micros)
It’s certainly been a reality since say 1930 that it would have been the way to go, if you could have easily bought a diversified basket of those stocks and held them (which we know wasn’t possible until the 80s).
Over-weighting them now means we expect that history will repeat (which in investment is never a sure thing).
Small companies, new companies, now have so much more private equity, at so much better terms than ever before, there is far less pressure on the next Apple to seek public funding.
A great example of this was Google. They secured so much private funding that when they went public they were immediately a large cap stock. Will we need to have another depression type event where money supply shrinks so rapidly and so much that there’s a new need to access public markets early in corporate development. That’s part of what data from the past predicts.
There’s another issue the experts talk about which is that large companies now are more aggressive than ever in acquiring smaller potentially disruptive companies. A great example of this is Instagram. A potential threat to Facebook’s dominance, they were bought for $1B. The valuation of Instagram now $100B. To get the rewards of Instagram you needed to be invested in the large caps.
Considering how few companies will create the value in these indexes, if this is the new trend, the place to get the rewards on small caps, might well be the S&P 500!
I agree with you regarding Int’ls. They’ve been an awful place for money because those nations have embraced thinking that capitates their growth and outlook. But it still seems to provide the movements opposite of the US. If someone said, go 2 fund, not 3 (ditching the Int’l) I don’t think I have a good counterargument.
Have no idea about the future. The data I quoted indicates that small caps returned 135.21% 2015-2024 which is better than the returns for the decade that preceded, 2005-2015.
I’ve done better with international stocks by buying individual companies rather than the index alternative, which I have never liked.
As for the U.S., Jamie Dimon, CEO of JPMorgan Chase was quoted in Bloomberg “If you were to take all of your money and put it in one country, it would still be America.”
Morningstar sees brighter prospects in Europe, but that may change next month!
With the current heightened level of uncertainty, there are differing opinions, but one consensus seems to be smaller returns on U.S.equities.
Chasing returns is not a good approach in my opinion.
I too don’t know the future
As I said at the top, I’m not making an argument against small caps, just that there’s a lot of thought amongst the pros these days about what the future of small caps (with regard to total return, and risk adjusted return) truly holds. This is a big change from say the 1990s when small cap returns weren’t being reconsidered.
I’ll add that it could well be that this is a great time to buy small caps, as if they’re a bit out of style, perhaps their are things brewing that will restore their former glory?
I’ll say this about the proposed portfolio. It is consistent with a backward look at long term returns. If you really believe that what’s been will happen again, then his portfolio makes a lot of sense.
Based on the added risk involved, I think it’s reasonable to assume that small-caps will outperform over the long haul. But will they outperform over the 20 or 30 years that someone is drawing down a retirement portfolio? That’s not at all certain, which is why I’m not sure a retiree should overweight small and micro-cap stocks.
That’s a very valid concern as well.
I guess when I read about how certain things are bound to happen (because they happened before), I think, didn’t the industry assure us for decades that active management had to outperform the indexes. Only in retrospect can we see how that notion was mathematically flawed. We only find out after the fact when they assumptions and dependence on history is mistaken.
I wonder about back-testing various portfolios, identifying which assets outperformed, then betting they will continue to do so in the future. For my domestic stocks, I think I’d be happy with a simple US total stock index fund, maybe with a small cap index fund added for “tilt”.
I had similar reactions. I have a small cap tilt but not that much and not that small. And a third in international. And too many holdings, almost all with capital gains. Oh well.
I think the most interesting thing about Bengen’s work is probably the withdrawal aspect more than the portfolio.
I actually really like Bengen’s work even though I don’t find it appealing to me as a strategy. It seems to me that it’s got tilt towards left tail fears and there’s a high likelihood of under-spending. (Especially if you’re thinking about a legacy goal).
But it’s a great framework for starting to think about withdrawals in retirement.
I think time horizon, asset allocation, whether you desire any residual portfolio value at the end of the horizon period ( for inheritances or a margin of safety) must be considered before deciding on how much to take each year.
The famous 4% Rule (or guideline as many prefer) and it’s many variations provide a predictable dollar amount each year, making it easy for future budgeting or financial planning.
One could opt to take a specific percentage of the portfolio annually, which will provide “lumpy” annual distributions. Of course, you must be flexible enough to be able to get by on less cash in some years.
Buffett has written a lot about investing, but virtually nothing, to my knowledge, about distribution of assets in retirement. From his discussions about annually donating shares, and about annually selling off a fraction of shares as a substitute for an annual dividend, I infer he would favor the “lumpy” option.
It seems to me that the first approach prioritizes the amount of cash withdrawn and smooth annual amounts, while the risk is premature portfolio depletion (depletion at the end of the 30th year is considered a success). The risk of the latter approach is living with volatile annual cash distributions but not the risk of portfolio depletion.
Maybe a hybrid of the two approaches?
There are different approaches, and I run several each spring, using December 31 data. I’ve come up with a range of possibilities which I think of as low and high boundaries. It is likely that reality lies somewhere within those bounds. One is the IRS RMD calculation for 2024 which was 4.37%. I took 5.01% but if I were to use all of the accounts it was 2.75%. That’s low if I use a base 4% plus inflation added each year of withdrawals.
Benz lays out a few approaches at Morningstar. One I don’t see mentioned very often is Bernicke’s Reality Retirement Plan. Ty Bernicke CFP wrote a paper in 2005. “Reality Retirement Planning: A New Paradigm for an Old Science”
He suggested that the withdrawal strategy should reflect a retiree’s spending declines throughout retirement. Bernicke suggests that a retirement spending plan should reflect this decline or reality. He goes on to suggest that an individual could retire sooner and spend more during the early years of retirement.
Bernicke used an early retirement model, beginning at 56. He assumes various spending decline rates from that age through age 75. The rates of decline are between 3% and 4%. At age 75, he switches to an inflation-adjusted amount until the end of the plan at age 85. The switch is to take into consideration medical expenses.
One problem I have is that the plan, and life, ends at age 85. Actuarial tables indicate this may be unrealistically short. The plan allows one to spend with an initial withdrawal rate of 8.7%!
I do like that Bernicke uses decreasing spending which is true for many retirees. However, who knows what the real spending will be for each of us? In my family’s case medical costs have increased, but my probable longevity has decreased significantly.
I’m not a proponent of the FIRE movement (Financial Independence, Retire Early). Primarily because I found planning for a 30 year retirement to be sufficiently difficult. Retiring at 55 means one could spend 40 years in retirement. That’s a planning bridge too far for me.
What if 55 +30 is a realistic outer expectation for the individual based on likely health, ancestral markers etc?
If there is one thing the FIRE movement is good at it’s helpful on focusing on the Work v Live equation to help individuals work out what they are willing to sacrifice to get FI and subsequently when and how they might want to RE. I’m not sure that outside of certain well remunerated professions there are slews of people hanging it up at 40 but as a means for more people to get to traditional early retirement ages having put in the personal psychological work it has merits.
Late life is an unknowable lottery for most of us. I’d rather flex for the go-go retired years at the expense of less financial featherbedding for no-go, on the basis that all plans need to be adaptible to be resilient.
I’m not a fan of FIRE per se, but like the idea of younger and middle aged folks striving to achieve financial security at a younger than retirement age, if possible. Not necessarily so that they can retire early, but so that they can explore numerous options, which can include working less hours, switching to a lower paid profession which they might enjoy more, or just keep on at their regular job if they like it.
I’m nervous about predicting my annual spending in the future. I imagine I will spend less in some areas, but non-covered medical expenses and long term care loom as possible high dollar items to pay for.
I still see this as choosing among strategies, some which will maximize cumulative dollars withdrawn (but at just how much risk to the portfolio?) and those which prioritize preservation of at least part of the portfolio beyond the time horizon period. Insurance companies have the advantage of risk pooling. I must keep in mind my wife’s family, even more than mine, has seen many live to nearly a hundred. Which for us is 28 years away. My main “ace in the hole” is that we are fortunate to have a relatively large portfolio in relation to our annual spending needs, hence we can get by comfortably with lower percentage withdrawals than many advocate.
I appreciate how thorough you are in researching what you post about.
Mr. Buffett has been criticized for not providing dividends to shareholders. I’ll paraphrase but he has been quoted “I can manage money better than the investors in BRK” and there is some truth to that.
That has certainly been true historically. In more recent years, Buffett has mentioned that if Berkshire paid an annual dividend because some shareholders desire it, this will present taxable income for many other shareholders who do not wish to receive one. His solution to those in the former group was to simply create a synthetic dividend (My words not his) by annually selling some percentage of their shares, such as 3.2% or something like that.
Seems like a resonable response.
One hybrid approach is the “guardrails” approach:
https://humbledollar.com/2023/03/riding-the-rails/
Thanks for the link.
Morningstar performs an annual starting withdrawl rate study and found the following:
https://humbledollar.com/eb623da6-ff2b-4ec5-9976-603efa5fb5b8
Their calculations are based on forward looking projections, rather than utilizing Monte Carlo’s historical returns.
That link isn’t working. I get a 404 Not Found.
I posted this link in response to a February 27 article
https://www.morningstar.com/retirement/how-retirees-can-determine-safe-withdrawal-rate-2025
The table suggesting initial percentage withdrawal rate adjusted annually for time periods in addition to 30 years is interesting for two reasons. One, even the shortest period, 10 years, had higher rates for those with equities in the 50% or so range compared with all equity, or 80, or 90%. And two, the rates quoted were associated with a 90% success rate. Which I read as a 10% failure rate risk.
Norman, thanks for the review. There’s lots of interesting writing about the SWR, but Bengen is the original, and worth a read. I’ll add this book to my list.
I’ve always considered that Bengen is a reasonable starting point. I run a Monte Carlo simulation which uses historical data since 1927. It adjusts spending each year for inflation. Using a 4% starting point the simulator indicates that there are 124 possible 30 year periods in the available data. A total of 6 cycles failed, for a success rate of 95.2%. Increasing to 4.7 initial percent indicates that 25 cycles failed, for a success rate of 79.8%.
Using Bernicke’s Reality Retirement Plan one reduces the inflation-adjusted yearly spending by 2-3% per year starting at age 56, and then stabilizes at age 76 to keep up with inflation. Using an ititial 4.7%, the Monte Carlo indicates a 100% success using that approach.
That’s all very well, unless you happen to be one of the failures. Not buying a 20.2% failure rate. That’s one person in five. Don’t understand the reference to reducing spending, thought we were establishing a sustainable rate.
I agree, I would not accept an 80% success rate for a 30 year period. That means that only 99 out of 124 30-year cycles succeeded. I’m sure we each have a success rate that we will accept. While we don’t want to run out of money in our lifetime we also may have limits to how much we want to leave for others. It is a balancing act.
I prefer to use these calculations as starting points which restart every year using the previous year end data as a new starting point. From my perspective these are dynamic calculations and are not set in stone. After 7 years of withdrawals my spouse and I have at most 22 years remaining. I can and do make period adjustments. When recalculating one can include current health and longevity, dollar amount of the portfolio, projected spending, market conditions, cash and bond returns, etc.
I re-run several calculators each spring using the December 31 data so as to determine the actual withdrawal for the following year. An assumption that is made is that all of the funds that are withdrawn will be spent each year. After seven years I’ve found that in my case that is not necessarily true. I’ve also determined that I have oversaved significantly.
I use Bengen’s caution about using the COLA method of increasing withdrawals each year because “Inflation causes you to raise your withdrawals and the higher the valuation of the stock market, probably the closer we are to a bear market.”
Past performance is no guarantee of future performance. Maybe if you had retired 30 years ago 5% would have been sustainable (although wouldn’t sequence of return have been a factor?), but that doesn’t mean returns will be as good going forward. I just finished Bernstein’s revised “Four Pillars of Investing” and if I remember right he was saying 3.5% was the highest safe rate.
I’ve read everything he has published. One gem in the revised edition is his discussion of how his views on certain topics have evolved since the first edition. And, I recommend to all who enjoy his writing to visit his website, http://www.efficientfrontier.com, which has been largely inactive for years, click on and read his one page essay “Of Viruses, Distressed Sales, and Stocks’ Rightful Owners”.
I’ve heard and read a lot of financial advice that suggests going beyond that 4% threshold, so it makes sense to me that he would update it. Given the past decade of stock market performance, you probably could have gotten away with much higher than 4%, maybe more like 6-7% and still come out with great growth in your portfolio. I always just think of 4% as an anchor point, and like a loose rule of thumb, not the end all be all of retirement withdrawal. It’s a simple tool, not a natural law. It’s all based on past performance as well, so it’s not impossible that future performance will render it outdated.
As a complete amateur, I’m not qualified to argue with Bengen. But having read many others discuss this topic, I’m skeptical one can start with 4.7% and then annually adjust for inflation for 30 years and not risk premature portfolio depletion. This is very different from starting with a higher number, such as 5% or so, but being prepared to lower the amount when market performance warrants it.
And the CAPE has been running well above historical levels for many years now. I think I’ll stick with around 3% or so to play it safe. I’m sure I’m being over-cautious, but I’d worry about exceeding 4%, unless one is prepared to forgo the annual inflation adjustments when markets are not performing well.