Wednesday, August 24, 2011

Capital Market Expectations, Asset Allocation, and Safe Withdrawal Rates: Additional Results

This is the online appendix for my article, Capital Market Expectations, Asset Allocation, and Safe Withdrawal Rates,”which is currently scheduled to appear in the January 2012 Journal of Financial Planning. First I will show Figures 3 through 9, which are described in the paper, but which are omitted from the final published version due to space constraints.  After showing these figures, I include a variety of figures which show the relationship between portfolio returns and volatilities for a wide variety of retirement durations and acceptable failure rates. An earlier draft of the paper which does include Figures 3-9 can be found  here


Supplemental Figures



  1. Wade –


    Your paper(s) and sets of graphs showing how SWR is related to its key variables are WONDERFUL! A visual gold mine to help people see the sensitivities and judge optimal SWR tradeoffs.

    Yes, on your graphs showing how various SWR failure rates are related to return-rate mean and StDev, your superimposing of the efficient frontier appears to me entirely valid – and creative. But your graphs have the greater value of enabling assessments and comparisons of other portfolios too, such as any two or a whole range of model portfolios. I’d suggest illustrating and emphasizing that. While the efficient frontier reflects great ideas, that curve itself is based on such absurdly indefensible precisions that it’s useful only as a guideline, say for creating and assessing a model portfolios set.

    You knew I’d love your graphs – to me they are, for SWR, a visual atlas of guidelines and tradeoffs for EVERY plan analogous to what Pathfinder can do for a SPECIFIC plan. (For a test, I entered specs for the 4.3% optimum in your example graph, and got 10% failure, illustrating complete agreement.)

    I think the GREATEST value of your graphs and underlying approach is to advance the investment planning focus from so-called “return” and “risk” for the individual year to the investment purpose, dollar value for the investor’s future needs and goals. For this, your graphs incorporate key factors omitted in the “risk-return” view – dollars, years, and the mighty effects of compounding along the way.


    After creating all those graphs, you deserve a day off. But I have some suggestions for further work, mainly to boost people’s appreciation of, use of, and benefit from your graphs.

    With so many graphs, a visitor may get an initial feeling of being flooded. There’s great potential for you to alleviate this feeling, and also illustrate the graphs’ value and uses, by doing more spotlighting of high-priority things the graphs show and ways to use them.

    Your superimposing an efficient frontier curve is a good example. As I mentioned above, I think you can illustrate more value showing use of the graphs to compare two portfolios, and even better to compare a range of model portfolios such as stocks/bonds in 10% allocation increments.

    As I went through your graphs, I was jolted by your graphs 8 and 9. Graph 8 shows that as you raise withdrawal rate above 4%, failure probability increases FAST! For the many tens of millions of folks who do not or will not have enough for dcent lives at 4%, that is terrible news. Terrible for the whole economy and country too. We need to do more to help those folks make best decisions.

    Then Graph 9 jolted me, with its illustration that as you take out %s for fees, SWR drops hard. According to the Investment Company Institute, less than 10% of people’s fund money is in index funds. That means fees. And there are various complicated tax situations. Maybe we need a second set of all your graphs with 2% or 3% taken out for “fees etc.” Those graphs would be very depressing, but perhaps more realistic.

    Dick Purcell

  2. Thanks a lot Dick. We've been discussing this by email. I definitely need to add some more descriptions here before the article is released in January. Best wishes, Wade

  3. Wade, I really enjoy your papers and have learned a lot. When are you going to tell us, assuming baseline/required spending needs are met with some fraction of savings/pension/social security, how can we maximize the real 30 year returns from the remaining "play" money that could come out in fits and spurts to buy toys and trips?
    John Walton

    1. Thanks John :)
      I know you're joking, but I'm not the person to ask about maximizing returns. I'm a simple index fund kind of guy. Let's hope Mr. Market is kind to us.