The 1% Difference: The Importance of Setting the Right Withdrawal Rate During Retirement

Let me start this post off by saying that this idea isn’t for everyone. Some retirees are comfortable spending down their assets during retirement and would therefore find this idea to be way too conservative. I’m not necessarily against spending down assets, but I do think it is important to exercise caution because future income is dependent upon the asset base (also called a capital base). The smaller the base, the smaller the income.

As long-time readers probably know, I have been tracking a few model portfolios over the last several years. With 2007 coming to a close, I have been updating the portfolios over the last couple of days. These portfolios consist solely of exchange-traded funds. Unfortunately, the exchange-traded funds in the portfolios have only been around since mid-2001 (some of the ETFs are even newer than that). Such a short history makes it difficult to really judge the prudence of a strategy. So,… this year I went back and assembled a hypothetical portfolio going back as far as I could using index data. I then organized the annual returns into the following table and then computed the annual returns of the portfolio based on the allocation. Here’s the annual results since 1992:

This particular portfolio is allocated 50-50 between stocks (both domestic and international) and bonds. I think this allocation is both conservative and worthy of consideration for a retirement portfolio. Notice the returns for the years beginning in 2004 are shaded grey. Those years represent the returns for the actual exchange-traded funds that represent those asset classes (the bond class is represented by equal parts in the iShares Lehman Brothers Aggregate Bond Index Fund (AGG) and the iBoxx $ Investment Grade Corporate Bond Fund (LQD)).

After I assembled the above table of returns, I was able to run scenarios using different withdrawal rates, assuming a beginning retirement balance of $1,000,000. The results were interesting. For example, take a look at what happens with a 5% withdrawal rate:

In order to make this hypothetical as realistic as possible, I assumed that the index returns had management expenses of .48%, which is equivalent of most of the iShares ETFs used in the portfolio. I also assumed portfolio management expenses, which include both a FOLIOfn advisor fee and an advisor fee. A do-it-yourselfer could definitely reduce those expenses if they wanted to handle their account on their own.

Anyway, notice how the initial withdrawal was $50,000 at the beginning of 1992 and $69,079 at the beginning of 2007 for an average annual increase of 2.18%*, which did not keep pace with inflation. Also note that the asset base grew from $1,000,000 to $1,411,799 at the close of 2007.

Now, what happens if we adjust the withdrawal rate to 4%?

Agreeing to withdraw a smaller amount each year, makes a sizeable difference over the years:

So, although the initial withdrawal was $10,000 less in 1992, it was only about $5,000 less at the beginning of 2007. Why is this so? Because a smaller withdrawal rate left more money to compound over the years. The retirement balance at the end of 2007 was $258,818 higher under the 4% withdrawal strategy.

What’s the point of all this?

I think this example shows that if you can afford to withdraw less from your retirement account, you should (at least in the beginning). You can always take out a greater percentage later if the need arises. On the other hand, if your goal is to spend down your assets during retirement, you can ignore this post. Just remember that once it’s gone, it’s gone.

*Calculated using the following formula:

[($69,079 ÷ $50,000)1/15] – 1

13 thoughts on “The 1% Difference: The Importance of Setting the Right Withdrawal Rate During Retirement”

  1. The way you are calculating withdrawals isn’t the typical way that is done. Normally a 4% withdrawal rate means you withdraw 4% the first year, and then adjust for inflation every year afterwards.

    It doesn’t really make sense to do it your way. For example note that if there are many bad years in a row and your portfolio drops by half so will your income. On the other hand, it will never drop to $0.

  2. Andy,

    Taking withdrawals can be done many ways. I don’t see how the way I illustrated “doesn’t make any sense.”

    My point was that in order to keep your income increasing, you must withdraw less than you are making in return.

    And, although it is possible to have several losing years in a row, the portfolio is pretty well diversified.

  3. I was always under the impression that a 4% withdrawal rate means you withdraw 4% of your principal each year, not that you withdraw only 4% the first year, then withdraw that same amount, adjusted for inflation.

    I think JLP’s method is correct, since the 4% number (as far as I know) comes from the assumption that retirement will be at least 25 years so the safe withdrawal rate is 100% / 25 = 4% of the principal.

  4. No. The correct interpretation of the 4% SWR is 4% of the initial balance increased by inflation annually. This does mean your balance is liable to grow to large figures, but those are necessary to overcome disasters like the great depression or the inflation ravages. In pleasanter times, a larger SWR is possible, but how sure are you this is or will be one of those times?

  5. [ “…if you can afford to withdraw less from your retirement account, you should (at least in the beginning)…” ]

    …seems very obvious that consuming less of one’s savings/investments — causes one to have more of it left, over any given period of time.

    Not much of a newsflash.

    However, one must also factor inflation into the equation.

    Real U.S. monetary inflation is now 8-10% annually — totally negating the average return on the example portfolio. It might well be wiser to cash-out of the entire portfolio … and invest in something that is less vulnerable to large inflationary losses.

  6. MorrisonH,

    The tone of your comment bugs me.

    The point of the post was to show how much a person’s withdrawal rate affects their retirement account balance.

    Although I do think inflation is higher than what the government is reporting, I don’t think it is anywhere near the 8-10% number you’re using. If your inflation numbers are correct then people should sell everything and run fer them thar hills!

  7. Official numbers for inflation are around 3% annually, inclusive of all sectors. A nice site for this is

    Though its early to tell, the inflation rate may be higher this year, just from eyeballing the monthly figures.

    In any case, JLP, I’ve seen a somewhat similar thing on Paul Merriman’s site, and he does it over a longer time period with the sort of inflation-adjusted withdrawals Andy suggests, as well as straight percentages at a couple of different levels. he comes, essentially, to the same point you make: if you can take less out, it can make what seems to be a disproportionate difference. it’s worth a look.

  8. Related topics:

    JLP, I’ve developed a model similar to what you’ve done in ordder to run all kinds of SWR senarios. What is shocking is how little changes in assumptions can yield massive changes in results – whether you die rich or dirt poor.

    I sometimes wonder if the typical 3% inflation figure is a reasonable guide for retirement planning assumptions. While certain things, such as food, fuel, and utilities, r.e. taxes, are bound to increase with inflation, what I’ve noticed about retirees is that they often tend to reduce personal expenditures over time, as they become less energetic and mobile. They don’t eat as much, don’t drive as much, don’t shop as much, etc. Also, if your housing cost is fixed (if you have no mortgage or carry a fixed rate mortgage), then another potential source of inflation is somewhat mitigated.

    Also, with respect to the oft-quoted 4% SWR, I could see adjusting this upwards, based on other non-financial assets (such as real estate) that could be monetized later in life. After all, if you make it to age 95, chances are good you won’t be staying in your own home (at least not alone) and even if you did, reverse-mortgage lenders would likely be lining up at your door.

    The other thing that seems to make a huge difference in my calc is taxes. I just don’t know enough to figure out what kind of tax issues I’ll face in retirement. Makes a massive difference in how much gross I’ll need to withdraw to meet projected expenses.

    Of course the biggest wildcard of all is healthcare – what will you need, how soon will you need it, and how will you pay for it. Maybe this alone is one good reason to err on the conservative side.

  9. I love this kind of stuff…

    According to Bernstein (Four Pillars of Investing) the withdrawals are supposed to be adjusted upward each year for inflation. However he also says that flexibility is the key to any withdrawal strategy so if you have a few bad years (especially at the beginning of your withdrawal phase) then you should reduce the withdrawals, so taking a straight percentage is not a bad strategy, as long as you have enough to survive on.

    I like the way you deal with commenters – take no s***!


  10. Mike (FourPillars),

    You said:

    “I like the way you deal with commenters – take no s***!”

    Normally I’m very nice and forgiving but sometimes the tone of a comment is just too much.

    Anyway, your comment about flexibility is right on! That’s why I think retirees need a budget and a plan so that they can have the flexibility to take out less if they don’t really need the money.

  11. The CPI inflation numbers have not seemed very realistic in years. While housing, gas, food, cars and insurance, etc have been increasing by nearly double digit rates the CPI is continually reported at 3%.

    Given the sharp drop in the dollar this past year its obvious that our buying power is much less. Whether the dollar drops or prices increase – it all equals rising inflation.

    Perhaps a better inflation rate would be found by looking at the annual growth rate of your own personal expenses.

  12. Miguel,

    I think it’s best to start on the low end and adjust withdrawals upward if and when necessary.

    That’s cool that you run scenarios. You’re a geek like me!

Comments are closed.