Sunday, April 8, 2012

Scott Burns: "How Much Will You Leave on the Table?"

This weekend Scott Burns wrote an excellent column ("How Much Will You Leave on the Table?") for the Dallas Morning News about the article "Spending Flexibility and Safe Withdrawal Rates" co-authored by myself with Michael Finke and Duncan Williams for the March 2012 Journal of Financial Planning. This is quite an honor for us, as my understanding of the history of these matters is that Scott Burns was the one who earlier brought the Trinity study into widespread prominence. Michael Kitces also took a look at our article at his blog last week, ("Utility Functions in Financial Planning - A New Framework For Decision Making?"), and Doug Nordman scooped them both with an excellent description of an earlier draft which he wrote last November ("Is the 4% withdrawal rate really safe?").

Scott's column prompted this comment at my blog entry about the article:

Wade -
Thanks for your colleagues and your for this research. As someone in the real world who will be retiring in the very near future I appreciate practical research and advice. I understand the need for theory but how does that get applied? Maybe Scott Burns over simplifies these ideas but people in the real world want to know how to apply something to their specific situation.
Thanks Matt Bly

It's a good question.

And it's an important one, because in particular as Doug Nordman describes, most of my earlier withdrawal rate research focuses on the concerns that I had and that I continue to have about the sustainability of the 4% withdrawal rate rule for retirees in recent years.

I understand that co-author Duncan Williams is working further to address this question with his research, but let me just provide some of my own comments for now about it.

The way I see it, the original purpose of Bill Bengen's research (that link will bring you up to speed on this issue) was to ask a very basic and important question: what is the highest withdrawal rate that can reasonably be used without running out of funds?

But a lot of subsequent research which should have been branching out became too stuck on this narrow question and ignored the fact that retires may have access to more income sources from outside their personal savings. I think an extreme version of this narrow focus can be seen in Rory Terry's research article from the May 2003 Journal of Financial Planning when he wrote:

This question clearly depends on the definition of an acceptable probability of failure and a definition of “acceptable” is certainly client specific. My opinion is that most investors would find failure rates in the five percent-and-higher ranges described in Ameriks, Bengen and Pye, for example, to be unacceptable. I believe that most investors would find even a one percent probability of failure to be excessively high when dealing with irreplaceable assets and considering the extreme costs of failure. (my emphasis)

For a lot of retirees, running out of wealth may not be as catastrophic as he makes it sound. There's also Social Security. For some people, there are employer pensions. Some people may have a fixed income annuity. Some people may have made arrangements with their children that the parents will use a more aggressive stock allocation in order to try and provide a larger inheritance with the understanding that this also increases the risk of depleting wealth in bad market conditions such that there is a small chance the children will need to help their parents out. And so on.

The problem with arbitrarily choosing an acceptable failure rate, as traditional safe withdrawal rate research would ask you to do, is that it leaves aside the issue that you may ultimately give up a lot of potential spending power by using a low enough withdrawal rate to satisfy a low acceptable failure rate.

I don't think Scott Burns oversimplified our article. There is more math and theory in the article, but ultimately that was to formalize the concept and demonstrate the tradeoffs with clear examples of well-defined retiree preferences.

I'm glad in particular that he emphasized our "percentage of retirement spent without wealth" measures as a contrast to traditional failure rate measures.

I think for the time being, the way to apply this theory in practice is to just keep some of these concepts in mind as you formulate your retirement income strategy:

1. Retirees may like to spend more, but there is diminishing additional enjoyment coming from more and more spending each year.

2. There is a tradeoff between spending more now and increasing the probability that you will be forced to spend less later.

3. Retirees need to consider carefully about the threat to their quality of life should they happen to run out of wealth at some point while still alive. For retirees with a solid income floor from Social Security and other sources, the flexibility about being able to run out of wealth may be sufficient to support a higher spending rate and higher potential failure rate. 

4. With a sufficient spending floor, those of us to tend to save a lot may find that it is reasonable to go ahead and splurge a bit without feeling guilty about it.


  1. Excellent points. How about placing the risk of longevity on the US government? By delaying social security benfits until 70 and using the pre-age 70 years to spend down assets instead of taking SS, many people can create a significant income floor. Many people start ss early because they have concerns about the government, stability of ss, and a host of other worries, some valid, some not. In the end, I believe using social security benefits as a significant part of your income floor can provide a lot of long-term security in a person's financial plan.

    1. That sounds like a pretty good plan to me!

  2. Thanks, Wade!

    Your analysis is very pertinent to U.S. military retirees. Many of them (in their 40s) can live a very frugal life from their pension alone (and Social Security later). That annuitized income lets them invest more aggressively, spend a higher withdrawal rate (especially if they cut back temporarily during down markets) and be much more confident of never running out of wealth.

    Unfortunately most military retirees feel obligated to work a bridge career to a traditional retirement age (and yet another pension). It's very hard for them to convince themselves otherwise, especially if a retirement calculator shows any success rate less than 100%. I think your analysis is the foundation of a new generation of retirement-planning software that will more accurately model retirement asset allocation and variable spending.

    Then I just have to lead the skeptics to my book & blog to help them figure out what to do all day...

    1. Thanks Doug. And thanks for the reminder too that about incorporating variable withdrawal rate strategies... at least flexible even before running out of wealth.

  3. Wade - Thanks for the followup. I think these discussions often focus exclusively on the investment side because that is where fees can be earned. (Nothing wrong with earned fees.)

    However when there are other sources (social security and pension) they take a large weight off the investement/savings side. And so there can be a lot of flexibility on that side.

    The initial Bengen article providing a great starting point. But it also seems we have barely gone beyond that.I think with the work on longevity and spending and flexibiity the indstry is starting to go beyond the basic concept.
    Thanks again - Matt