Insights on Using the 4% Withdrawal Rule From Its Creator

Bill Bengen explains why he now suggests a 4.5% withdrawal rate and what the biggest threat to his withdrawal strategy is.

Charles Rotblut leads a class in AAII's new Essential Investing Video Course. Go to https://www.aaii.com/ves for more information and to subscribe.

Article Highlights
The safe withdrawal rate is the percentage that a retiree who had the unfortunate luck to experience both high inflation and poor market conditions could have withdrawn and not run out of money.

A starting point for determining what inflation rate should be used for increasing withdrawals is the consumer price index (CPI). The CPI should be adjusted based on personal circumstances, such as known fixed costs.

The biggest threat to the 4% (now 4.5%) rule is not a period of low returns, but rather a lasting period of high inflation, with prices incurring percentage increases approaching the double-digit range.

William “Bill” Bengen is a retired financial planning practitioner. We spoke about his groundbreaking research into what is known as the 4% rule (now 4.5%) including its backstory and what investors using it today should take into consideration.

Charles Rotblut (CR): Your research is based on the concept of a safe withdrawal rate. Could you explain for our members who are unfamiliar with the term what a safe withdrawal rate is?

William Bengen (WB): It’s really important to recognize that the word “safe” should be taken with a grain of salt since it’s based upon what’s happened historically. If markets behave differently than they have in the past, what was safe in the past may not be safe in the future.

But in any event, the research I conducted basically went back to look at and reconstruct the experience of investors from 1926 through 1986, retiring them every quarter. I gave them a portfolio of diversified investments. Then I just used the actual investment returns and the actual inflation rates they would have experienced.

I looked for the worst case, meaning periods when the portfolio ran out of money the soonest. That’s what the 4.5% safe withdrawal is based on: that unlucky investor who retired in October of 1968 and ran into some terrible stock markets and terrible inflation. The combination just devastated their retirement portfolio. By taking out inflation-adjusted 4.5% withdrawals, their money just barely lasted 30 years. [See the box below.]

CR: Originally, in your 1994 study, “Determining Withdrawal Rates Using Historical Data,” (Journal of Financial Planning, October 1994), you used a 4% withdrawal rate. What prompted you to increase the withdrawal rate to 4.5%?

WB: I included more asset classes.

Originally, I only worked with two asset classes. I used U.S. large-company stocks and U.S. intermediate-term government bonds. I then added small-cap stocks. The small-cap stocks added enough of a boost in terms of return to allow the withdrawal rate to be increased.

It was originally around 4.2%, actually. Including small-cap stocks raised it a little bit to about 4.5%. This shows you the importance of having a diversified portfolio during retirement.

Historical Worst-Case End-of-Year Portfolio Values

A person who retired in October 1968 had the unfortunate luck of encountering periods of both very high inflation and terrible market performance. The combination led to what Bengen found to be to the worst 30-year period for taking retirement withdrawals, as shown in the chart below. This is the period he based the 4.5% rule on.

The dollar amounts for each year are the year-end portfolio balances. An allocation of 35% large-cap stocks, 20% small-cap stocks and 45% intermediate-term bonds was used. Taxes were excluded from the calculation. As you can see, the retiree ran out of money in year 30 by using a 4.5% withdrawal rate.

In discussing the chart with us via email, Bengen commented, “It appears, during the middle of retirement, that after the initial trouble, everything was hunky-dory. Alas, not so. After about the 20th year the portfolio nosedived and never recovered. This is typical of ‘failing’ portfolios from which too much is being withdrawn. The seeds of destruction are sown early in retirement, in this case by a major bear market plus sustained high inflation.”

Figure 1.

 

CR: Before discussing the subject of portfolio allocation further, I’d like to ask what prompted you to do the research on withdrawal rates in the first place.

WB: I became a financial adviser in the late 1980s. I was a baby boomer and most of my clients were baby boomers. By the early 1990s, I had a lot of clients who were 20 years away from retirement. For the first time, they were seriously thinking about how much they should save for retirement, how their investments should be set up during retirement and how much they could afford to withdraw.

I started to look into that information for them, but I couldn’t find anything anywhere pertaining to that. I looked through all my CFP (certified financial planner) manuals and I looked through all the financial planning material. I couldn’t find anything because—quite frankly—at that time retirement was basically a 10- or 15-year affair. People didn’t have to worry about having large amounts of money saved because they weren’t going to live long enough to enjoy it.

A lot of people in my generation expected to live into their 90s and were concerned about having money for 30 or 35 years. So, I decided since I couldn’t find information anywhere I’d do the research myself. I just got the data on investments and the CPI (consumer price index). I then set up the spreadsheet and started running the calculations. I had no idea where it would end up 25 years later.

CR: What was the original reaction to your analysis and findings? It’s pretty well accepted now, but what was the reception when you first had your study published?

WB: It was two-sided. I actually got hate mail from some people who refused to believe that only 4% was all you could take in withdrawals. A lot of people had been advising clients for years and telling them they could take up to 6% or 7%, even 8% based on some very simple, straight-line calculation. But they weren’t taking into account the effect that a major bear market has on your portfolio. It’s just devastating for a retiree. So, there was that.

There were also people who looked at what I did and said “Hey, this is cool. Let’s start using it.” Over time, the reaction became more and more positive and I guess you could say probably even more accepted, as it is today.

CR: Interesting. I didn’t, and I don’t think many of our members, know the strategy’s backstory. Could you explain how the strategy works for those who are not familiar with what it asks retirees to do?

WB: The mechanics of doing the withdrawals?

CR: Yes.

WB: You pick a number based on your preference or maybe a consultation with your financial adviser. Let’s say you take a 4.5% withdrawal and you have a $1 million portfolio. So, the first year you’re going to apply that 4.5% against $1 million. You will take the $45,000 ($1,000,000 × 4.5% = $45,000). That will be your first year’s withdrawal. The second year, you throw away that 4.5% withdrawal and never look at it again.

All you do is look at what consumer inflation was the year before and add that percentage to your initial $45,000 withdrawal. So, let’s say inflation was 10% the year before. You then have to add 10%, or $4,500, of withdrawals for the second year, so your second-year withdrawal totals $49,500. Each year you just increase the withdrawals with inflation. Basically, your lifestyle is keeping pace with the inflation rate.

CR: What would you recommend somebody use for an inflation rate?

WB: Well, in my research, I liked it when you were actually determining your personal inflation rate. But you could use the CPI, even though you’re not obligated to do so. Individuals each have their own different inflation rates depending upon what their expenses are.

Let’s say a person in retirement has a substantial mortgage payment that’s fixed. That portion of his expenses is not going to grow with inflation; it will be fixed. So, he could potentially use a lower inflation rate, a personal inflation rate. Some people might need a higher one. But in general, I would start clients out with the assumption that they would be using the CPI as a starting point for discussion.

CR: Just to be clear, start with the CPI and then modify it based on your personal situation?

WB: Yes.

CR: Getting back to allocations, I know you said that when you went to 4.5%, you added in small-cap stocks to the portfolio. On your recent Reddit thread—which, by the way was great—you suggested allocating between 45% and 55% to stocks, 35% to 45% to bonds and 10% to cash. Is that a correct way of summing it up?

[Bengen participated in a question and answer session on the social media website Reddit in September 2017; you can read through the session here.]

WB: Yes, I said that’s really good.

It’s pretty close to what Harry Markowitz, the great Nobel Prize winner, uses. I asked him about how he personally allocates his portfolio. He said that he’s using the 50% stock/50% bond set up, so that allocation has some pretty good credentials behind it.

CR: In terms of stocks, how are you splitting it? Do you still suggest large caps and small caps?

WB: Well, I would think most investors are probably using a much wider array of investments. They will have some international equities, they may have some real estate. I didn’t use all those asset classes in my research because, until recently, I couldn’t find databases that went as far back as the 1920s. Now that they’ve become available, I’ve updated my research to include them.

Retirement investors should have a very well-diversified portfolio spread out among a number of asset classes. The small-cap stock is kind of like a proxy in my research to represent all the other asset classes that retirees would normally employ in building a portfolio.

CR: What about on the bond side, particularly in the currently still-low interest rate environment? Should retirees just be thinking about intermediate-term government and high-quality corporate bonds? Or should they think about diversifying on the bond side as well?

WB: The research I did indicated that intermediate-term bonds performed the best. You didn’t gain anything by using long-term bonds at all because of the increased volatility. In fact, when I did my research, you could eliminate your bond portfolio completely and replace it with cash. Because of the 0% volatility of cash, you would have done totally well without bonds. I wouldn’t recommend that today because cash is paying so little. We’re living in a very distorted environment today. It’s hard to figure out what to do. But I think investors should probably just follow some basic diversification principles and have different types of bonds.

Folks who hold intermediate-term bonds shouldn’t go particularly long with their bond maturities, but rather should consider having some short-term, some international, some emerging markets … to build a diversified portfolio. This will actually perform the best for an investor.

CR: On the cash side, you talked about holding the 10% in cash. I believe you said that the cash allocation comes out of the bond portion and its purpose is to provide a buffer for retirees. Is that a correct way of summarizing it?

WB: We’re always going to need some cash for withdrawals. I assume retirees will be withdrawing on a regular or monthly basis, so they want to have some cash in their portfolio to handle withdrawals.

I think it’s good to have a little extra, maybe a year or two of withdrawals in cash just in case you run into a bad stock market, a bear market. I find it very comforting for retirees to know they’ve got 10% or 15% cash and they’re not going to have to sell their stock investments in a really bad market environment. They can just live off of that cash for a few years until things recover.

CR: In terms of taking withdrawals, if market conditions are good instead of bad, should retirees think about taking, say, half of the withdrawals out of stock and half of it out of bonds? Or should they split the withdrawals, either equally among all of their asset classes or proportionately?

WB: When I did my research, I assumed we were rebalancing once a year and therefore the cash would come out of whatever asset class had done the best.

As a practical matter, I also looked into rebalancing less frequently. I found out that if you rebalance once every six years, you can actually add about a quarter of a percentage point to your withdrawal rate. The reason is that stocks tend to run in long bull markets and if you rebalance too frequently, you’ll cut those gains off.

So, effectively, the money comes from wherever you’re rebalancing from. You just want to make sure that if you rebalance your stocks, your bonds and your cash, the money should naturally flow from how the investment performed over the previous 12 months.

CR: In terms of taxes, I’m presuming your research treats the withdrawal rate as being independent of tax considerations. Obviously, at the time you did your original study, there weren’t Roth IRAs, but their existence now does not alter the withdrawal strategy, correct? The withdrawals should be calculated based on the investor’s starting portfolio value regardless of the type of account(s) the portfolio is held in.

WB: That’s correct. When we’re dealing with a tax-preferred portfolio like an IRA or a Roth IRA, and so forth, the 4.5% rule applies. The net withdrawals for taxable portfolios are probably about 10% to 15% less. This is based on when I did those calculations; the actual reduction depends upon the prevailing tax rate.

CR: If the retiree has both a traditional IRA and a Roth IRA, should they use the same withdrawal percentage for both? Should they try to adjust in a certain way? Or should they go with the simplest approach, say, “here’s my 4.5% starting point and then I’ll adjust it going forward?”

WB: It shouldn’t make any difference as long as they’re withdrawing from a tax-advantaged account.

When you start getting into a lot of different accounts and types, you almost need software to sort that out and make sure that you’re making provisions for the taxes because that’s an expense. Part of the expenses you will incur are your taxes. That’s where having different types of accounts can get complicated.

CR: What about required minimum distributions (RMDs)? At some point, the absolute dollar amount of the RMDs will exceed the dollar amount suggested by the calculated withdrawal percentages.

WB: They almost certainly will later in life. Basically, an RMD is a transfer of dollars from one account to another, with part of it going to pay taxes. So, it doesn’t really affect things too much. But, once again, it probably requires software to develop a long-term plan and take the taxes into account that are being expended when money is still taken out.

CR: Once retirees take an RMD, I presume they should set aside any excess beyond the amount determined by the inflation-adjusted withdrawal rate—say, put the excess into a savings account or into a taxable brokerage account since in the future the money will be needed to help fund withdrawals. Is that the thought? Any excess amount taken out attributable to the RMD is not money to be spent. It’s money that they should set aside.

WB: It depends on the account size. That money could just flow into a taxable account and be allocated among the same investments and the same asset classes they are currently using right away. I don’t think you want to have an excessive amount of cash sitting around and not earning decent returns.

CR: As far as withdrawal rates—and I’m sure you are aware of this—there are some people now saying we should use lower withdrawal rates if future returns turn out to be lower than historical averages. I’ve seen calls for withdrawal rates of 3.5% or lower. Do you have any thoughts about this?

WB: We’ve had periods of time in the last 90 years when investment returns have been quite low. And yet the 4.5% rule applied. My greatest concern is not the returns going forward.

My concern is having a spurt of very high inflation that would turn out to be lasting. That would tend to jack up your annual withdrawals on a permanent basis and you’d really start running through your portfolio. I haven’t seen any evidence of this occurring, so I don’t have any concerns about it happening in the foreseeable future.

[Editor’s note: Since the 4.5% rule adjusts withdrawals up each year by the rate of inflation, high levels of inflation would increase the withdrawals by a much larger amount in absolute dollars than low levels of inflation would. Since the inflation adjustment is calculated based on the prior year’s withdrawal amount, the effect of high inflation would be compounded and potentially reduce your savings at a much faster rate.]

Even if we earn 1% or 2% for the next 10 years on a balanced portfolio, which I’ve heard some sources quote as being probable, it will just mean that people are unlucky to retire at this time. They will run their portfolios down during retirement, but they won’t necessarily exhaust their savings completely. People are going to have to get adjusted to the point of view that if they start with $1 million, 10 or 15 years from now they’re going to have a lot less. They may still have enough to fund their retirement, but it’s not going to be as much fun as it is for many others who were able to build their wealth up during retirement. It’s a little scary.

CR: With inflation, are we talking about something like we saw in the 1970s? Is it that level versus, say, 4% or 5% inflation that worries you?

WB: Yes, something getting close to double digits for decades—that would be very difficult to deal with. Then I would be concerned about the 4.5% rule holding up, especially if it’s combined with a really big bear market. But so far, I haven’t seen that level of inflation on the horizon. Have you?

CR: No, fortunately I haven’t.

WB: It doesn’t mean it won’t come. It could come as a surprise, and suddenly. But not yet.

CR: What about a bad sequence of returns? Say somebody had the unfortunate luck of retiring in 2007 or somebody is getting near retirement and they start seeing a bear market. Should they stick with a 4.5% withdrawal rate as long as we don’t have that extended period of terrible markets and terrible inflation?

WB: Yes. I’ve done a lot of looking at January 2000 retirees because they’re one of the few groups that retired and faced two major bear markets in their first decade of retirement. Those investors are still reasonably secure with that 4.5% rule, surprisingly. It’s probably because they’ve benefited from an enormous recovery of having held stocks over the last nine years. Now, if we get a third bear market in the next year or two, all bets are off. But so far the 2000 retirees and the 2007 retirees seem to be holding up reasonably well.

Once again, as you know, markets go down. That’s part of the process. And they recover, that’s also part of the process. That’s why I’m not too worried about it.

I’m much more worried about inflation because inflation will cause you to increase withdrawals. Those increases will get locked in for the entire remainder of retirement. There’s no recovery to the portfolio then.

We’ll see. You know, they say this time is not different, but this whole environment feels a lot different than I can ever recall in my career.

CR: It does.

What about a flexible approach to taking withdrawals? I’ve seen some advisers suggest it. Vanguard has even published a study about using a floor-and-ceiling approach to calculating the withdrawals. What is your opinion on this?

WB: When I wrote my book [“Conserving Client Portfolios During Retirement,” (FPA Press, 2006)], I looked at that as a potential methodology, another withdrawal scheme that you might use. The stock market fluctuates, however. Are you familiar with Michael Kitces’ work?

CR: Yes, I’ve spoken with Kitces and we’ve published articles co-authored by him in our magazine.

WB: A couple of years ago, he developed a terrific chart where he plotted market valuations against the safe withdrawal rate year by year. It was an amazingly close negative correlation between the two. The higher that stock valuations are, the lower the safe withdrawal rate turned out to be. [See the box below.]

Valuation Ratios and Safe Withdrawal Rates

The chart below compares the cyclically adjusted price-earnings ratio (CAPE), a rolling 10-year valuation indicator developed by Yale professor Robert Shiller and the maximum safe withdrawal rate (“SAFEMAX”) a retiree can take. SAFEMAX is based on the performance of inflation and investment assets in the subsequent 30 years.

As financial planner Michael Kitces originally showed (“Resolving the Paradox: Is the Safe Withdrawal Rate Sometimes Too Safe?,” The Kitces Report, May 2008), there is an inverse relationship between the long-term valuation of the stock market and how much retirees can withdraw without running out of money. During periods of high valuations, lower withdrawal rates are warranted, but during periods of low valuations, higher withdrawal rates can be sustained. A challenge for investors is not knowing what actual future market returns will be. Bengen notes that his SAFEMAX computation does not involve stock market valuations although, as the chart demonstrates, the two have historically been very strongly negatively correlated in the past.

Bengen used a portfolio allocation of 35% large-cap stocks, 20% small-cap stocks and 45% intermediate-term bonds. Taxes are excluded from the calculation. The SAFEMAX data only runs through 1987 to reflect actual 30-year withdrawal periods.

Figure 2.

 

His conclusion was that when you get a CAPE (cyclically adjusted price-earnings ratio) above 20, you should stick with the lowest, the safe, withdrawal rate because otherwise it’s too risky. We’re certainly well above that now. So, I don’t think any kind of a scheme where you attempt to try to take out 5% or 5.5% now is likely to work.

I expect, at some point, that there’s going to be another serious decline back to more normal valuations. You’re going to have to start scaling back what you withdraw each year. It might be painful, after you have misled yourself about the kind of lifestyle you really think you can afford.

CR: Some other people have suggested other strategies. For instance, Wade Pfau, who’s done lots of work with Kitces, has written about using reverse mortgages to supplement withdrawals. As I’m sure you are aware, there is also a camp that says investors should just annuitize all of the money needed to cover expected fixed expenses at retirement. Any thoughts about these alternative strategies?

WB: They’re worth looking at if people are concerned that their money isn’t going to last. I think people ought to sit down with a competent individual who has their interests at heart and work through the numbers. Doing that may make sense.

Bonus Audio

Click here to hear audio of Bengen’s suggestions for those nearing retirement and the biggest lesson he’s learned over the course of his career.

Discussion

David Van Knapp from NY posted over 8 years ago:

I would be interested in knowing Mr. Bengen's view on the role of dividends in his retirement strategies. Many retirees use dividend growth strategies to fund all or part of their retirement. Every dollar generated organically by the stocks (or funds) in a portfolio reduces by a dollar the assets that need to be sold to achieve any withdrawal rate. With this in mind, some retirees build their portfolio strategies around dividend growth stocks. Inflation adjustments happen "automatically." AAII offers a paid service based on dividend investing. I'd be interested in hearing Mr. Bengen's views. I am a longtime fan of his work.


William Bengen from CA posted over 8 years ago:

Dear David, Thanks for your interest in my research. I haven't spent a lot of time in my research on dividends; I have focused on "total returns", or the sum of dividends and capital gains. As long as the total returns you earn from your dividend-paying stocks are comparable to those for the asset classes I use in my research (eg, S&P 500), my research would be applicable. Having said that, dividend-paying stocks make sense to me, as they provide a continuing flow of cash to replenish your money market fund, which is the source of withdrawals during retirement. This can reduce the need for selling investments to rebalance an account, which might give rise to a taxable event. Best regards, Bill Bengen


John Schuerman from MS posted over 8 years ago:

We continue to see diversification recommendations around 55% equities and 35% bonds and 10% cash. In the current environment it seems we leave a lot of earning power on the table with low bond rate. I personally feel a high percentage of dividend paying equities is a better option. I recognize the increased risk; however, most bear markets don't last more than a few years if you retain a couple years cash to help get through the recovery you would be better served. Most of the dividend paying equities increase in valve 15-20% this year. That rate of increase and the dividend return rate of 2-3% makes it to get into Bonds.


William Bengen from CA posted over 8 years ago:

In response to John Schuerman: Thanks for commenting on my article. It is very tempting, in this odd time of low bond interest rates and higher stock dividends, to substitute dividend-paying stocks for bonds in the portfolio. However, I caution you that by so doing, you place your retirement portfolio at risk. My research is based on a study of 90 years of data, including the period of 1926 through 1966, when stock yields were much higher than today, and were consistently higher than bond yields. Yet the evidence shows that, even during those periods, replacing bonds with stocks tended to reduce the safe withdrawal rate. The effects of a major bear market on a portfolio heavy in stocks is frightful to see. Originally, I had developed a 4.1% withdrawal rate using only large-cap stocks and US Treasury intermediate-term bonds. When I did the research for my 2006 book, I replaced a portion of the large-cap stocks, which paid good dividends, with small-cap stocks, which paid no dividends at all. Yet the safe withdrawal rate increased to 4.5%. It appears to me that "total return" from a balanced portfolio of stocks and bonds trumps higher income from a stock-heavy portfolio. Please be careful with making long-term changes to your strategy based on short-term phenomena. Best regards, Bill Bengen


Alan Ruger from WA posted over 8 years ago:

Extremely useful information for those planning for retirement. I would greatly appreciate your thoughts on how the 4.5 withdrawal rate should be adjusted for someone wishing to plan for a shorter retirement. For example somone in their 70’s, planning for a 24 year retirement.


William Bengen from CA posted over 8 years ago:

In response to Alan Ruger: Alan, thanks for your interest in my work. In my book, I provide a table listing "safe" withdrawal rates for different time horizons. For 25 years, "SAFEMAX" is 4.7%; for 20 years, it is 5.2%. Note that the optimum equity allocation also varies with time horizon. Best regards, Bill Bengen


Tom Canfield from PA posted over 8 years ago:

How do you reconcile your findings with the results we can find at FIRECalc and the Vanguard Nest Egg Calculator? For 4.5% spending rate and the investment mix you suggest, both find the probabilty of depleting a portfolio is 15% to 20% at the 30 year mark. Both show the point of first depletion at roughly 20 years.


Bill T. from CO posted over 8 years ago:

In Wade Pfau's new book "How much can I spend in retirement", he briefly mentions adjusting stock allocations from a normal 50/50% to 25% when the Shiller CAPE is 1/3rd above its mean to 75% when it is 1/3rd below its mean. [P159, Valuation Based asset Allocation] This timing model allows a higher withdrawal rate from 3.93 to 4.58%. This makes sense to me, though recently it would have cost some appreciation on an otherwise higher allocation. Thoughts on that approach?


Bill T. from CO posted over 8 years ago:

Original article link cited in prior comment re Wade Pfau valuation based allocation in which he mentions you as the idea's originator https://mpra.ub.uni-muenchen.de/35329/1/MPRA_paper_35329.pdf


William Bengen from CA posted over 8 years ago:

In response to Tom Canfield: Tom, thanks for your excellent question. I am not familiar with FIRECalc's method of computing withdrawal rates, but I did look up on the Vanguard Website their Next Egg Calculator. It showed for a 50% stock/30% bond/20% cash allocation, a 4.5% initial withdrawal rate fails 17% of the time. This agrees with your assertion above. Note that Vanguard employs Monte Carlo simulation, which produces a 30-year investment record by randomly selecting, for each year, actual returns from a historical database. In order to achieve 95% confidence that savings will last 30 years with the above asset allocation, the Vanguard calculator requires a withdrawal rate of only 3.6% In contrast, my methods use actual historical returns and inflation rates in the order in which they occurred. Vanguard's methods create sequences of returns and inflation which probably never happened in reality. As a result, they may generate "worst case" scenarios worse than anything that has ever happened, while my methods search for the worst case that has actually occurred. Note also that Vanguard uses different asset classes than I do in my research. My incorporation of small-cap stocks significantly raised the safe withdrawal rate in my research, from 4.1% to 4.5%. Vanguard appears to use returns broad market indices, which are dominated primarily by large-cap stocks. Is Vanguard's approach more accurate than mine? I honestly don't know. I am not sure anyone can be certain. One thing I am certain of is that neither my method, nor Vanguard's, can predict the future. Waiting out there somewhere might be a "worst than worse case" scenario which may invalidate any "safe" withdrawal rate predicted by either method. Let me add that I am a great admirer of Vanguard and their effort to serve investors well with low-cost, well-managed funds. I use their funds in my personal portfolio. But our approach to computing "safe" withdrawal rates, as you can see from the above, is quite different. Best regards, Bill Bengen


William Bengen from CA posted over 8 years ago:

In response to Bill T: Thanks for your note. I am a great fan of Wade Pfau and read everything he writes about retirement income. I have not studied in great detail the correlation between Shiller CAPE and withdrawal rate. However, Michael Kitces, a celebrated financial planner who has also done some important research in the area of withdrawal rates, produced an interesting chart some years ago. It showed a strong negative correlation between CAPE and each year's safe withdrawal rates. So, I am not surprised at the conclusion that Wade reached, although I believe he may have been the first to quantify the "boost" to SAFEMAX by using a timing strategy. However, for those who wait for a low enough CAPE to invest fully in stocks, it has been a frustrating 25 years, as CAPE has been below its average of about 16 only about 25% of the time during that span (if that much). Today, of course, CAPE stands at more than twice its long-term average. It will be interesting to see if it does indeed "mean revert", and even drop below its long term average. The only time that has happened in the last 25 years was for a few weeks in 2009. Investing ain't easy. Best regards, Bill Bengen


Lance Casual from PA posted over 8 years ago:

“I’d like to follow up on the point about FIRECalc. I’ve used it extensively and love it. It builds sequences of return the same way you do (clearly stated on its Home Page). But it’s conclusions for your suggested mix are COMPLETELY different. It would say a portfolio will deplete in 20 years, not 30 years, for a 4.5% spending rate. Something's amiss between the two of you.” Can you please take a look at Firecalc and test your conclusions with their tool? A detailed explanation of the methodology is included in the tool. Thanks. - Lance


Bill Bengen from CA posted over 8 years ago:

In response to Lance Casual: Lance, I always find it difficult to compare my methods to others. It takes a long time to understand other methodologies, their assumptions, etc. But I have looked at the FIRECalc set briefly and here are my thoughts: FIRECalc states in its "Your portfolio" section: "If you leave this section alone, FIRECalc assumes your retirement portfolio is invested in a "couch potato" portfolio of 75% stock index and 25% bond funds, with a 0.18% fee to the fund." In my research, I found that a 75% stock allocation leads to suboptimal withdrawal rates. I use 50% to 55%. This assumption alone could easily account for the difference in outcomes. Also, FIRECalc uses a different data base than mine; theirs extends back to 1871, mine to only 1926 (although I use quarterly retirement dates, and they appear to use annual.) I couldn't identify which lines on the graph corresponded to which years, so I can't determine if their results are due to inclusion of data prior to 1926. Also, does FIRECalc assume a tax-deferred portfolio? I couldn't determine that from their website, although I could easily have missed it. All in all, I am loathe to comment on a pretty neat-looking tool like FIRECalc which probably took many hours to build and maintain. Kudos to the authors. But in a limited time of study, I can't really offer much explanation for the differences between mine and their approach, other than what I said above. I have a lot of confidence in my approach, given its admitted limitations. Best regards, Bill Bengen


Bill T from CO posted over 8 years ago:

Bill, Thanks for the response. You are correct. Retirement investing is challenging! I don't suspect that the CAPE ratio is not the full answer. The yield curve may add insight along with measures of euphoria/ boredom/ panic. That said, the market will often try to disappoint the most logical investor.


Jack Kulpa from WI posted over 8 years ago:

I have read Mr. Bengen's work at length, especially his CONSERVING CLIENT PORTFOLIOS DURING RETIREMENT, which jibes with Jim Otar's UNVEILING THE RETIREMENT MYTH. Both works rely on historical--not Monte Carlo--data, and both writers arrive at the same conclusion; i.e., at withdrawal rates of 4% or less from a balanced portfolio, you will never run out of funds; however, at withdrawal rates above 6% you are sure to deplete your portfolio in less than 20 years. Thank you, Mr. Bengen, for taking the mystery out of creating and depending upon a non-depleting portfolio.


William Bengen from CA posted over 8 years ago:

In response to Jack Kulpa, Jack, thanks for the nice words- and thanks for buying my book! In a few years I hope to have an updated version published, which incorporates the research results I have developed over the last twelve years. Best regards, Bill Bengen


Dave Gilmer from WA posted over 8 years ago:

Bill, Thoroughly enjoyed your article. I have long thought with a little better asset mix 4.5% is quite doing looking at past history. What do you think about combining this with a variable withdrawal strategy such as the one based on the IRS RMD tables? I have found this to improve your overall retirement spending if you can just be satisfied for a short few bad years with slightly less. The good years more than make up for it. I do agree that real data is the better way to go. Monte Carlo simulators basically introduce a lot of data that would never appear in real life. The reason for the above is hidden in your comment about inflation - that inflation is the bigger driver of why you may run out of money. I have studied quite a bit the reasons investors ran out of money and it was not always the great depression that did it, because during that time expenses were actually deflating rather than inflating. The reason inflation is so deadly is that it hardly ever "resets" your expenses by going negative. I do believe dividends are sometimes overstressed and I doubt they will change the results of your research, which is based on how much I can withdraw before I run out of money. The answer to this is based almost entirely on proper asset allocation or withdrawal method and the total return of the investments you hold combined with the return sequence and inflation (or deflation.) I have also found that a method such as AAII founder James Cloonan's "Investing at Level3" which I have modified for my own use (and talked about in another comment thread) has been quite useful. So thanks Bill and James for all you have contributed to my success!


Roderick Johnson from WA posted over 8 years ago:

In retirement planning consideration is not given to retirement plans that may provide all or most of your retirement needs. It also doesn’t consider how much debt/mortgage that you may have. Tax rates change because of continuous changes in tax law. We are in our 70s have no debt of any kind and have secure retirement income that pays more than our requirements with enough for occasional traveling. We have tax deferred investments, and regular taxable investments. Our withdrawal rate is zero, with the exception of RMD. I think that retirement planning is totally incomplete to simplistically say have this or that allocation and withdraw at a certain rate!


Roderick Johnson from WA posted over 8 years ago:

In retirement planning consideration is not given to retirement plans that may provide all or most of your retirement needs. It also doesn’t consider how much debt/mortgage that you may have. Tax rates change because of continuous changes in tax law. We are in our 70s have no debt of any kind and have secure retirement income that pays more than our requirements with enough for occasional traveling. We have tax deferred investments, and regular taxable investments. Our withdrawal rate is zero, with the exception of RMD. I think that retirement planning is totally incomplete to simplistically say have this or that allocation and withdraw at a certain rate!


Roderick Johnson from WA posted over 8 years ago:

A key that was left out in the previous post was we have always lived below our means (frugal) not on credit cards. And we were fortunate enough to have defined benefit plan and participated in deferred compensation plans.


William Mcginty from GA posted over 8 years ago:

Your article and the comments are interesting and continue to point up that each person's retirement and withdrawal rate are very individual. I am retired for 18 years and a fan of "Die Broke" (authors Stephan Pollan and Mark Levine) and do not want to leave any residual to Charity, Church or Descendants. If RMD of my IRA covers my living and extensive travel expenses what should my asset allocation be?


John Strudwick from FL posted over 8 years ago:

I have enjoyed the article and your research. What would be the impact on the withdraw rate during an extended period of deflation such as that being experienced in Japan?


William Bengen from CA posted over 8 years ago:

Response to Roderick Johnson: Thanks for your comment. I believe you must consider yourself extremely fortunate in having adequate income from pensions and other sources to fund all your retirement needs. Most people must rely on at least a portion of their retirement savings to pay their living expenses during retirement. It is true, my research does not take into account mortgages, pensions, and other financial features. That is because I was focused on a single issue: making a portfolio last for a specified number of years. I think of the retirement portfolio as a "cow" that is going to provide milk during retirement. In effect, all my research is centered on the behavior, care and feeding of the cow, so as make the cow a reliable giver of milk during the rest of one's lifetime. The issue of maximizing one's withdrawals from one's retirement portfolio is of interest to a great many people. Others have done excellent research in the areas you mention, and any financial advisor worth his or her salt would certainly take into account all the elements of one's retirement finances, not just the retirement portfolio.


William Bengen from CA posted over 8 years ago:

Response to John Strudwick: John, thanks for your comment and question about deflation. A deflationary environment would probably be extremely friendly to retirees. As you may recall, the early years of the Great Depression were deflationary, and despite the huge stock market decline during 1929-1932, retirees survived with a 5.5% withdrawal rate because their portfolio withdrawals declined each year. One wonders why central banks, then, are working so hard to raise the inflation rate. As the saying goes, "Be careful what you wish for....." Best regards, Bill Bengen


William Bengen from CA posted over 8 years ago:

Response to Dave Gilmer: Dave, thanks for your comments and question. You asked about my opinion of a variable withdrawal strategy. Sorry if I get this wrong, but I infer that you are suggesting that retirees withdraw less during years of big portfolio declines, to help preserve the value of the portfolio. I was a little confused by your reference to the RMD tables, as they specify a continually rising withdrawal percentage. Quite frankly, I think it is natural for retirees to spend a bit less during periods when stocks are getting hammered, such as 2008. I think that's fine, from both a portfolio and a psychological perspective. We all spend money we don't really have to, and during tough times we can cut back. There is no "mandatory" in the 4% rule. Best regards, Bill Bengen


William Bengen from CA posted over 8 years ago:

Response to William McGinty, Thanks for your comments and question. You ask what your portfolio allocation should be in an IRA when the RMD covers all your required expenses. My answer is: I don't know! You have added an interesting new constraint I have not studied in my research, that of a continually rising current withdrawal rate. I will add it to my list of areas to investigate. Best regards, Bill Bengen


John Belanger from OH posted over 8 years ago:

Two points: 1. A portfolio composed of stocks that raise dividends at a rate greater than inflation will never be depleted if you only spend the dividends. 2. If one retires in a low cost of living area and has no debt it is possible live reasonably well on 2 social security payments. With a part time job you can do even better.


John Jones from NJ posted over 8 years ago:

I think your analysis is interesting and insightful. However, I would like to understand how you factor in income tax rates into your 4.5% rule. I have both an after tax investment account and a pre-tax retirement account. Both accounts combined constitute my overall savings. I assume the 4.5% rate would apply to my pre-tax retirement account. Should I use a different percent withdrawal rate for my after tax investment account?


Robert G from MI posted over 5 years ago:

There have been some analyses on gold in a portfolio; namely Ray Dalio and Golden Butterfly portfolios as an element to provide some stability long term. Any thoughts on gold and what percentage makes sense? thanks


JACK K from CA posted over 5 years ago:

I have read this 4% and now 4.5% withdrawal rule for years. The analysis assumes a static portfolio and never considers a portfolio growing 4, 6, 8 or even 10% a year and compounding growth. A static approach is probably a guaranty for running out of money earlier than later.


You need to log in as a registered AAII user before commenting.
Create an account

Log In