Finding the Right Withdrawal Rate: One Key to Portfolio Sustainability

by Maria Crawford Scott

Finding The Right Withdrawal Rate: One Key To
Portfolio Sustainability Splash image

Like all investors, individuals who are living off of their retirement savings need a blueprint to construct a successful portfolio.

But the blueprint that works for the typical investor who is saving for retirement won’t work for those who are living off of their retirement savings.

That’s because the goals of these two types of investors are not quite the same.

...To continue reading this article you must be registered with AAII.

Gain exclusive access to this article and all of the member benefits and investment education AAII offers.
JOIN TODAY for just $29.
Register for FREE
to read this article and receive access to future articles.

Log in
Already registered with AAII? Login to read the rest of this article.
Maria Crawford Scott is the former editor of the AAII Journal.


Bernard Biltek from Florida posted about 1 year ago:

This discussion does not consider that currently bond yields are at historical lows. Will the average bond yields be much lower over the next 30 years than they were over the 1926-1995 time periods. If so that will have a major effect on using this data to determine withdrawal rates.

H Padden from New Jersey posted about 1 year ago:

So, you must pay taxes from your 4% and you (very luckily) make 15% on your investments the first year. Your combined tax bracket is, say, 40%. You must, therefore pay 6% (15% X 40%) from your 4%. So, you are required to live on -2% in your luck year. Who will pay you that money?

You simply cannot use this simple rule inflexibly, but only as a very rough guideline for your overall planning.

Robert Brooks from Arizona posted about 1 year ago:

30 year averages are a bit unrealistic. For the first time since 1920, there is a 10 year period (1998-2008) in which investments in the S&P index would have produced a net loss!
Most of us over 65 are concerned that another precipitous drop in the market would essentially wipe our our stock investments in that a 50% decline would require a 100% increase to just break even.
It is very difficult to maintain a "relatively safe" return of over 4% in order to be able to withdraw at that same rate.
I have decided to anuitize a good portion of saving into a single life, no beneficiary, annuity (actually, diverifying among the top rated insurance companies). I can receive a 7.8% return of investment and principal for life. If there is money left over in early retirement, then those dollars can be reinvested into "riskier" assets.
The key, as you pointed out, is to maintain enought fluidity for emergencies and potential investments if the market really takes another dive.
We can thank Mr. Greenspan and his guru, Ayn Rand, for starting this lowest interest rate paradigm which has lasted for over 10 years now. It has had a major negative effect on retirement incomes and all of the retirement portfolios of major agencies (i.e. STRS).
I believe interest rates will rise again (cycles tend to repeat) and hopefully I will have enough cash to reinvest at a higher rate. Now is the time not to invest in bond funds as the upside potential is minimal and the downside risk is horrendous.

Bob Brooks
Prescott, AZ

James Stewart from Massachusetts posted about 1 year ago:

Seems that you would only pay 40% taxes on any money you withdraw from IRA type accounts, or 4%X40%=1.6%, for example, plus taxes on dividends, capital gains,other income,etc.

Robert Krisch from Pennsylvania posted about 1 year ago:

An alternative approach bases the withdrawal rate for each upcoming year on the current value of the portfolio at the beginning of the year, along with the retiree's life expectancy at that same time. This gives a substantially more irregular rate of annual return than the Bengen model that you outline, but has NO risk of running out of money before you die and takes automatic account of fluctuations in the year to year value of the portfolio. This observation is not a revelation. This is the basic principle on which the German retirement system and numerous other large governmental and private retirement systems work. It is also the dominant principle behind the US Social Security System, with various political tinkering around the edges to decrease the short term risk to the annuitant, but with no guarantee that these modifications are permanent.
I understand that this approach to withdrawal from retirement accounts has never been popular with financial planners because it has never been popular with their clients, who generally want a fixed lifelong stream of inflation protected income. I have always believed that with some effort to carefully explain this approach to people in parallel, the risk benefit tradeoff compared to the accepted Bengen approach would look much more attractive. Running out of money at an advanced age without any remaining earning power seems to me to be a much more serious risk than is a significantly fluctuating annual income. As a bonus, this approach also automatically eliminates the long term risk of taking out too little money if the investment markets do unexpectedly well over an extended period. As another bonus, it is very easy for intelligent individuals to apply to their own retirement accounts. Life expectancy tables are widely available on the internet, including multiple tables from the Social Security Administration.
In summary, I feel strongly that this general model should always be offered in parallel as a serious alternative approach to retirement withdrawal, both by financial planners in individual sessions and in articles such as this one, whenever the Bengen approach is presented as a "reasonable" one and that the Bengen approach should never be presented without serious discussion of alternatives such as this.

Robert Krisch from Pennsylvania posted about 1 year ago:

ADDENDUM to my post above. I apologize, but I somehow forgot to include an important, familiar and classic example of the alternative withdrawal plan discussed above. It is a fairly pure example and is widely used by many prosperous retirees in the US, including I am certain, many readers of this Journal. It is the minimum required annual withdrawal plan laid out by the Federal Government for retirees over 70.5 who own IRAs and/or any other lump sum tax sheltered retirement accounts. With these accounts, you generally don't even have to figure out the required withdrawal amount each year, since the managing institution will almost always do it for you, mail your checks to you regularly and do all of the required record keeping. The life expectancies used are generally for two joint lives, giving a lower than necessary withdrawal rate for single people, but you are always allowed to take out more than this minimum.

Steve Laube from Georgia posted about 1 year ago:

The classic 4-5% adjusted for inflation approach to withdrawal is flawed in several ways:computr modelsthat run "10,000" simulations usingrandomly selected returns remove the bjsiness cycle from the data, bastardizing the data to the point that one can say it is invalid. More importantly, these secnarios typically assume ALL withdrawals come from the stock portion of the portfolio, even though the asset allocation contains typically 50% bonds and cash. Why use this "safe" allocation and then not use it correctly to withdraw from the non stock portion when the market/portfolio is not at a high water mark. Selliong stocks when the market is down 30 or more per cent results in a double whammy wher ein more tock must be sold to raise money andthen that stock is gone forever when the recovery comes. Last, for many of us, the 3% inflation assumption is way overstated. Unless you are living a minimal lifestyle, age will take care of the gym membership, country club membership, travel and entertanment expenses, mortgage payment, big home/utility bills/taxes, and even the transportation as one goes from two or more cars to one and then none. Also, it is very easy in real life to use a combination wherein the 4% guideline is combined with the ability to cut spending via "substitution" and denial for a year or two if the portfolio is trashed and you are concerned; the withdrawal models assume no adjustment to withdrawals, even as one sees the approaching freight train of running out of money.

Richard Abbott from Florida posted about 1 year ago:


Michael Frey from Ohio posted about 1 year ago:

No way can one take out 4% (in real terms) and expect to maintain the value of the account over time. It's a pipe dream. I've seen this flawed argument over and over from people who should know better.

John Borrows from Ohio posted about 1 year ago:

Too much attention to what I cannot control and not enough to what I can.

During my 69 yrs. I've lived high on the hog and eaten pigs feet. Can't say that there was a lot of difference in the happiness and satisfaction at either end.

So, the first step is to decide on a minimally acceptable life style, and its annual price. Using whatever conservative estimate you want to decide if you can afford it. If you can't -- resume' time. But you probably can with quite a bit of cushion. Proceed to live that way and watch for truly valuable things to spend the excess on. If all is well with the reserves (money not needed to maintain the minimum acceptable) spend for it and enjoy. If things are going south with your reserves, add the thing to your bucket list.

What is different here is that it seems if you calculate your income first with the idea that you will then do your budget to match, you'll find a way to budget to the income.

I see too many people who spend hours of work and lots of worry on maximizing or estimating their incomes without having a clue of how much they are spending on stuff that really isn't all that important to them. Until you figure out what you really want, you won't know anything about what you have.

Back to managing income, what I would really like to see is a plan for what to withdraw. Building up these assets I had this wonderful plan of steady, equal, investments leading to dollar cost averaging. An assurance that I was buying more at the lows and less at the highs.

The financial planners have not provided me with any way to cash in my investments that is better than a constant percentage rate assuring long term average recovery. If I take out a set amount each year (as suggested here) I get dollar cost average in reverse, assuring below average returns in the long run. I would not want to try to make my living selling that "plan" to paying clients.

Robert Franzen from Missouri posted about 1 year ago:

What I recommend is tracking your draw rate on a monthly basis using the following process:

1. Create a paper account and credit it with 4% (divided by 12) of your 12 month average portfolio value

2. Reduce the account by your actual draw for the month

3. Monitor the account balance and make adjustments as desired or required

The benefit to using a 12 month average is that values do not suddenly change radically due to market volatility.

If the account builds a surplus value over time, you may decide to reward yourself with an unplanned vacation. If the account builds a deficit value over time, you know it is necessary to reduce spending.

The two most important things one can do to reduce the risk of running out of money are:

1. Setup a disciplined process to manage and monitor as described above

2. Maintain a diversified portfolio containing as many asset classes as you feel comfortable with and rebalance regularly.

James Johnston from Nevada posted about 1 year ago:

The more I read about this subject the more confused I get.This is my first year in retirement, (lost my business) I'm 61.
Is there a good program that would help me determine a withdrawel amount showing diferent examples?.

Robert Krisch from Pennsylvania posted about 1 year ago:

In response to Mr. Johnston's post, I agree that this is a somewhat confusing subject, but hopefully the following comments will be of some use to you. My two adjoining posts from yesterday briefly discussed the easy, do it yourself program of treating your accumulated retirement assets as if they were in an IRA and calculating your regular annual withdrawals as if they were your required minimum withdrawals from the IRA. These are calculated very simply. The only information you need is the current market value of the accumulated assets and your current life expectancy or the joint current life expectancy of yourself and your wife. You then divide the market value by your life expectancy and that gives you the amount to be withdrawn for the upcoming year. It can be withdrawn either as a lump sum or in monthly or quarterly installments. At the beginning of the next year you repeat this procedure with the updated value of your accumulation and your updated life expectancy. Individual and joint life expectancy tables are readily available on the internet. For example, the IRS provides several such tables with their instructions for IRAs. Your age is below the starting age of 70.5 for mandatory IRA withdrawals, but I believe that they have data for younger ages for single lives but not joint lives. Those could easily be found at other websites or in a reference library. For more detailed information and for examples of typical scenarios, you just need to look at the section on withdrawals in a few books on IRAs. There are many such books in every public library. Remember, some major advantages of this approach are: 1) You can't ever run out of money if you follow theses rules, although as with any approach where you can't run out of money, if your assets continually produce a low rate of return your annual withdrawals will eventually become lower. As I described in the earlier posts this method is officially endorsed by the US government for IRAs and by a number of other governments for corresponding retirement accounts. 2) It is pretty easy to do it yourself, with no fees and a minimum investment of time and energy, as described above and detailed in numerous other sources on IRAs. 3) It works as advertised, without any fiddling, no matter how you have invested the accumulation, although as noted above, the size of the annual withdrawals will eventually depend on the rate of return. Accordingly, this approach does not alter the need for diversification between and within asset classes or the need to keep a large fraction of the assets in equities. 4) Unlike the Bengen 4% approach, if your investments do unexpectedly well for any extended period, this approach automatically adjusts your annual income upward without any effort on your part.
The major disadvantage of this approach is that your income will not be certain from year to year because it will be calculated from the total value of your accumulation at the beginning of that year, which may fluctuate a little or a lot, depending on how you have it invested and on what the markets did during the previous year. This would be less of a negative once you start collecting additional stable, inflation protected income from social security, but it seems to me that even in the absence of social security, the likely degree of unavoidable annual fluctuation would be a small price to pay for eliminating even a small probability (say 5% as in many of the Bengen type models) of entirely running out of money at an age when you are unable to work.

Alan Thelen from Michigan posted about 1 year ago:

Because of the impact and difficulty of predicting the sequence of (good) market year returns I have been considering allocating at least the percentage recommended for bonds to an index annuity. I would then forgo any bond investments with my remaining funds. Any thoughts, comments, or dirty looks?

William Porter from Illinois posted about 1 year ago:

I've been retired "officially" only since the beginning of February, but I think the actual situation for most retirees is a whole lot more complicated than simplistic rules, such as a "4% initial rate" or the IRA distribution rule suggest. Some us lucky few do get to collect traditional pensions, which others might need to create for themselves by purchasing an annuity from an insurance company (or companies!). This income, plus whatever Social Security pays you, provides a guaranteed minimum annual income -- which is something most retirees desire most of all. But if you have other investments in stocks, bonds, whatever, inside or outside of IRAs, 401k plans, etc., in addition to the amount set aside for annuitization, a more flexible approach can be adopted for those additional funds. Yes, your income will fluctuate if you use the IRA withdrawal rate as a guide, but if this is in addition to a fixed annuity-style income (and not instead of a fixed, annuity-style income) derived from Social Security, traditional pension or an actual annuity, then such fluctuation should be tolerable. Indeed, with a proper mix of investments, the IRA-withdrawal model can provide a growing income stream, part of which can be invested outside of tax-deferred accounts, to continue to grow in net asset value, so that the fraction invested in equities becomes ever larger as you age.

It's important to look at ALL of your investable funds, whether in an IRA, 401k plan, or taxable accounts, and calculate the maximum allowable withdrawal based on this total. If you have too much in tax-derferred accounts, and are forced to withdraw more than this calculated value to satisfy IRS requirements, re-invest the difference in a taxable account.

Warren Flick from Georgia posted about 1 year ago:

For Mr. James Johnston, and others: at laterlivingblog dot com, I advocate a 4% rule, but without automatic inflation adjustments. Withdraw 4% of your initial total investments in the first year, then again in the second, then again in the third. Over that period, read about other approaches, and experiment with a calculator or a computer and spreadsheet (if you program one to model your investments). If your diversified portfolio is heading south, withdraw less. If your portfolio is bounding upward, cautiously increase your withdrawal by an inflation adjustment—maybe to 4.1% of your initial total investments—then hold that steady for some years.

Another factor not yet discussed is the asset allocation of the portfolio. In studies, authors often use the S&P 500 index along with a long series of bond returns, such as long-term, high-grade corporate bonds, or U.S. Government Bonds. Long data series are available for that data.

But keep in mind that a long-term data series is not a real investment. At laterlivingblog dot com, I’m writing about a real portfolio of four assets: domestic stocks, international stocks, bonds, and REITs. I use Vanguard index funds and trace portfolio performance over the most recent 11-year period, from the end of 2000 through 2011.

It turns out that a well-rounded portfolio having approximately 50% exposure to bond or bond-like assets and 50% exposure to stock and stock-like assets performed reasonably well over the 11-year period. I used a 4% rule—withdraw 4% of the initial value of the portfolio, then keep that amount constant. I also report results for an inflation-adjusted withdrawal, using 2.5% as the inflation rate. The approach is simple, and the results show that it would have worked well, even during the turbulence of the past 11 years. Of course no one can guarantee future performance. You might take a look at the blog posts—there are several about investments and withdrawals.

This coming Tuesday I’ll write about putting a human face on managing your investments, such as dealing with rebounding children or other relatives who need help.

Kenneth Dillman from Colorado posted about 1 year ago:

I've been retired for 20 years running a portfolio of 60% equities and 40% fixed income. My expenses are covered by social security, a small pension and withdrawal from savings. The savings are in a tax defered IRA, ROTH IRA and taxable brokerage account. I have more in savings now than when I started 20 years ago even with the 2007-2008 major downturn. If your going to manage your own portfolio and withdrawal rates you need to understand tax law, tax brackets and how RMD's are calculated. Everyone is different and one size does not fit all. Most retirement plans tell you to withdraw from taxable accounts first before using tax deferred accounts. This may keep your tax rate low in the beginning but your RMD may increase the rate later in life. What worked for me was to control my spending (stick to a budget)and take more out of my tax deferred accounts when the market was up but only enough to keep me in the same tax bracket. This reduces your RMD when you reach 701/2 and increases your taxable account that you can withdraw from and not increase your tax rate. Again, any standard withdrawal rate maybe a starting point but will probably not benefit most people. There is no substitute for knowing the tax law, having a deversified portfolio and controlling spending.

Clifford Wyble from Pennsylvania posted about 1 year ago:

Warren - Sounds like the best advise to date.

Dave Gilmer from Washington posted about 1 year ago:

Alan, stay away from the indexed annuities, until you can read and understand all the fine print. Once you do you will see there are much better options you can do yourself.

I like your approach of keeping the 4% constant for a few years and then adjusting.
Also like your term "bond like" investments. For me 2 pensions and one SS are all bond like income streams so I really don't need any other "real" bonds, especially in this interest rate environment. For my income I use a stock dividend portfolio, which has it's own built in growth of income. Since my "yield to cost" is of main concern, if any one of the dividends goes down, I just replace it with my large pool of money from my bucket 2 growth pool. Right now there are more companies with 25+ years of dividend growth than I even need.

Ted Nicholas from California posted about 1 year ago:

There are a number of articles on withdrawl rates during retirement but nearly all are too complex for quick review or too simplistic to be accurate. AAII has a chart in their Investor Classroom under Step 3 of Fund Mechanics investing and redeeming article. The X axis has the % annual withdrawl rate of capital (from 5 to 16%) and the Y axis has the % annual asset growth (from 1 to 12 %). You simply match the two variables and their intersection tells you how many years your savings will last. Here is the link:
I'm surprised at two things:
1. that the chart isn't at least referenced in every general article like this one.
2. that the chart starts at 5% withdrawl rate rather than the ubiquitously recommended 4% (or less).
It is without doubt the best single chart I've seen on managing retirement savings. Why? finances can change literally overnight. A quick glance at the chart with the new figure allows instant determination of a new withdrawl rate and/or time frame.

obie ephyhm from Minnesota posted about 1 year ago:

man, i wish the investment world would stop pumping this bilge out year after year just because it was the paradigm 45 years ago.

it doesn't work today. people live longer, save less, work longer, have had their savings eaten by inflation . . . on and on and on.

if you're smart enough to invest your own money, you should be smart enough to build a dividend income growth portfolio and then live within your means. no need for bonds. no need for stupid annuities.

articles like this are making me question why i became a life-time member of aaii. this is just standard financial "advice" designed to move products, part seniors from their money without providing any real security and it does not deal with the real world.

Curt Peterson from Georgia posted about 1 year ago:

I like Warren Flick's approach. Ted Nicholas' recommendation also is valuable. Unfortunately, market performance during the last decade and currently, doesn't mimic past performance cited in many of these withdrawal studies, and there is no way to know with any certainty that future peroformance will behave similarly to past performance. Monitoring spending and making sure your portfolio includes opportunity for growth of principal over time are two ways to insure that you do not run out of funds. This approach requires discipline and courage during Bear markets and other unexpected downturns.

Dennis Roubal from Michigan posted about 1 year ago:

I have subscribed to the 4% rule, with corrections, for a long time. Ultimate success requires some reasonable investment returns to maintain purchasing power. Most withdrawal programs are set up using historical returns as a guide. Future returns may be a real problem for income maintenance.
Research looking at past financial crises, like we are experiencing, indicate a very long period of low interest rates and under-performing equities.
Rheinhart and Rogoff and several other studies show results similar to Japan's current malaise. Their stock market peaked in 1989 and is about 80% lower. Their Treasuries are at about 1.6% yield.
If the U.S. follows this same path, and I think we will, a 4% withdrawal rate may be difficult to maintain.

J Morlock from New Jersey posted about 1 year ago:

I would like to thank to everyone for sharing their insightful comments.

Here is an article worth reading if you are looking for other retirement income planning tools and ideas.

Jim Otars's website is one of the best I have found for retirement income planning considerations.

I also like the consumption smoothing approach incorporated into the ESP financial planning software.

I would love to see AAII publish interviews with the owners of these two websites.

David Brenner from Alaska posted about 1 year ago:

I agree with Mr. Roubal that a 4% withdrawl rate is hard to count on in the long-term downtrend we find ourselves. A year-by-year approach of 4-5% max understanding that such may mean decreasing gross withdrawl amounts and a committment to try for a lower 3-4% goal may make a big difference going forward, especially considering the tandem of European problems and poor growth in the US. Hopefully the eventual recovery, when it arrives, will signal a return to the likelihood of success with the fixed (x% + inflation) withdrawl strategies discussed. As always, asset allocation will matter!

Charles Hinners from Wisconsin posted about 1 year ago:

The discussion below is limited to vanilla income annuities in which an insurance company pays a fixed amount for life in return for a lump sum now.
The central problem of financial planning is converting a lump sum to an inflation adjusted lifetime income

Nearly all annuities carry sales commissions of 3 to 5% that translate to a proportional decrease in benefit.
Annuities are priced of bond yields with durations matching life expectancy.

Berkshire Hathaway Life offers a no commission annuity in a limited number of states.
Their crediting rates for income annuities is 2.55% today 8/3/12.

It pays to shop carriers. offers lower commission alternatives.

They deal with several companies and more information can be gained by talking to Drew Hueler at 866-297-9835

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

Log In