Wade Pfau Answers Questions From May 18 A4A Webinar
Created: Saturday, 19 May 2012 22:53
Q: Comment on idea of drawing down assets on purpose anticipating "spend
rate" decreasing. Also comment on how Long-term care impacts.
The basic 4% rule indicates that you use a constant inflation-adjusted withdrawal amount throughout the entire 30 year period. But I have recently explored two reasons why people might decrease spending as they age. First, as I investigated at Advisor Perspectives in " How Do Spending Needs Evolve During Retirement?", people may naturally spend less as they slow down at later ages. Second, it is rational to plan to intentionally reduce spending since the probability of survival decreases with age. Both of these links explain how this allows for more spending earlier in retirement. Long-term care is the great unknown, though. The first link addresses that. How to properly plan for large and uncertain health expenses is an important issue in need of more research.
Q: With most respectiable estimates for stock and bond returns for the
next 7-15 years being in the low single digits maybe mid at best net of
inflation its painfully obvious 4% won't work. How does your research
account for projected returns on various asset classes which
historically over long periods of time is fairly easy to estimate?
Usually, research is based on the historical data. In the presentation, I did discuss two more alternatives considered in Joseph Tomlinson's work. These are to adjust bond returns to match current bond yields, and to assume a lower forward-looking equity premium. These assumptions do cause a substantial increase for the failure of the 4% rule as shown in the presentation.
Q: Great presentation! Doesn't this make a case for equity-indexed
products with income riders? I've never used them before but have been
considering as of late.
I will return to this question again later. I need to study equity indexed annuities more before I can provide a suitable answer.
Q: in terms of a retirement saver's asset allocation, does it ever make
sense to consider the present value of their social security payments as
a "government bond" in their overall retirement asset allocation?
Yes, I think so. Retirement income strategies should be based on a retiree's human, social, and financial capital, not just financial capital. So you can think of Social Security as sort of like a bond. Alternatively, in matching assets to liabilities, Social Security provides a good match for basic needs, which allows for a more aggressive asset allocation to meet discretionary needs. This is another way of developing the same sort of point.
Q: i follow wade's blog, and based on his research and his summaries of
other advisors' research, it seems like (at the end of the day), a
withdrawal rate in the 3-4% range is usually the "answer." have we seen
enough research to generally agree that range is "okay"?
3-4% is in the range of what is supposed to be safe in a worst-case scenario. It is not the average outcome. I have come to think that the floor/upside approach has a lot of potential for retirement. 3-4% should typically work out, but there is no guarantee and it is important to at least have your basic needs met, no matter the outcome. After all, as RMA curriculum board member Mike Zwecher wrote, retirees "only get one whack at the cat."
Q: In a previous webinar-I believe last year-you had a speaker
discussing MPT and highly diversified strategies whose returns would
have more than provided for a 4% SWP, even starting in adverse
environments. Isn't this therefore too simplistic and doesn't take into
account the ability of advisors and managers to outperform eoither on an
absolute or on a risk-adjusted basis.
This research assumes that people earn the returns provided by the market indices, and my Monte Carlo simulations assume a symmetric sort of distribution for market returns. If you develop a strategy that supports higher returns, lower volatility, or less downside risk than the underlying indices, then your abilities can be used to support a higher withdrawal rate. If you do think you can provide this superior outperformance, I do suggest that you take up Bob Seawright's Active Management Challenge.
Q: Guaranteed Annuities with Living Benefits solves most of the issues
Yes, they are designed to provide downside protection and upside potential all in one convenient package. Given their potentially high fees, I wonder if keeping these goals separate (such as with partial annuitization with a SPIA for a floor and a diversified portfolio for the upside) may still be a better approach for retirees. But you do raise an interesting point.
Q: Wade - Review Moshe Milefshiy's maretials
Moshe Milevsky is great. I once wrote that he is The Simpsons of the retirement planning research world. There is a famous episode of South Park in which Butters tries to bring disarray to the town of South Park. But for every evil scheme he considers, it turns out that The Simpsons already had a plot point about the same thing. In the case of retirement planning, Moshe Milevsky has already done it all. I did review his book, Pensionize Your Nestegg (Part 1 and Part 2). He has some new books coming out this year, and I will be sure to review those too.
Q: How does the intergration of Alternative Assets figure in to your
Q: Thanks Wade! very interesting perspective and analysis. I like the
use of Monte Carlo to show clients how their spending might turn out.
Q: Do I unerstand then that teh client needs to plan to have
guarantees through pensions and annuities , but inflation reduces
value by 50% over 30 years. and discretionary returns may not have
Yes, inflation risk is one of the big three risks retirees face, in addition to longevity risk (outliving assets) and sequence of returns risk. Inflation-adjusted SPIAs for flooring and building a TIPS ladder are two ways to deal with this.
Q: but what if the goal is to use up teh wealth in life. If I live 30
years , if Ilive 40 years, how much can I take from my accounts each
year using averages and having nothing left ?
In general, withdrawing a constant inflation-adjusted amount over retirement is not an optimal strategy. If you do not want to annuitize and do not worry about leaving a bequest, you could use a strategy such as withdrawing ( 1 / remaining life expectancy ) each year. This could cause your withdrawals to fall to uncomfortably low levels if we experience bad market returns, but it will help to make sure your wealth is getting used up.
Q: What are the ending wealth values for your MC alternative to the
Q: then you include admin fees and advisor fees and a retiree can not
take any money out of the portfolio?!
Yes, this is a problem. You can see about the impact of fees (which more generally represents underperformance from the indices... if managers produce alpha to counteract their fees, then this issue doesn't apply) in Figure 9 on page 17 of my article "Capital Market Expectations, Asset Allocation, and Safe Withdrawal Rates." Just follow the links for a free PDF download of the article.
Q: Have you considered the effects of a "diversified" portfolio over
these time periods? Does it make a difference? Small quibble, but I
would have to disagree that Social Security is guaranteed.
Social Security will need to have some cuts at some point, so your point is well taken. Nothing is truly guaranteed.
Q: Talk more about bond laddders as a hedge.
Q: Don't TIPs lack deflation protection?
You can't lose the real value of your principal if there is deflation. I'm not sure what you are getting at with the question. Are there other assets that will perform better with deflation? Maybe, and deflation could cause TIPS prices to fall for those who want to sell before the maturity date. But the idea is to hold the TIPS to their maturity date and not try to actively trade them. In this regard, you don't have to worry about deflation.
Q: How would you measure the maximization ultilization of assets on an
individual client basis? Asset optimization strategy measurements?
I'm sorry, I'm not sure, but I think your question might be getting at: how do we define the proper utility function for individual clients? It is hard. But as I explain in "Spending Amounts vs. Spending Value," it is impossible to avoid the idea of utility and just doing something which incorporates diminishing returns from additional spending may do a lot to help us get better solutions, even if the exact preferences of the individual cannot be modeled. I think this is becoming an active area for research in financial planning now with work by Joseph Tomlinson and by students and professors at Texas Tech University. Dick Purcell, who hates the over-mathematization of utility maximization, did suggest that the figures I showed in this blog entry do provide the type of info retirees could use to figure out their preferences.
Q: What did you mean when you said that using corporate bonds, as in the
Trinity study, caused more failueres? Is that because of behavioral
issues -- investors selling due to volatility?
No, it is not because of behavioral issues. The research assumes that people do not make mistakes, and that they stick to the chosen asset allocation. The issue is that corporate bonds are more volatile than intermediate-term government bonds, and this volatility causes slight increases in failure rates. Safe withdrawal rates are a delicate balance between return and volatility, and in this case the extra return from the corporate bonds was not enough to compensate for the extra volatility.
Q: What's you opinion about using scenario planning to comes up with return
and risk inputs on asset clases? Any system for economic scenario
analysis that you like? Have you designed one? Windham Capital's Mark
Kritzman came up with one and spoke about it at one of my webinars. It's
an interesting approach for making assumptions about the future.
(Kritzman manags about $35B and wrote the quant portin of the CFA exam.
Let me know if I can make an introduction to get you a look at his
This sounds interesting and potentially useful. Yes, I'd be happy to learn more. Thanks.
Q: One really important thing you said was that asset allocation is not
that important for retirees. Please explain. And try to include some
idea of what that means for advisors.
In terms of downside protection, there are a wide range of stock allocations that end up doing just about as well as one another. That showed up in Figure 2.3 and Table 2.3 of my presentation. In terms of upside, more stocks provides more upside potential.
What might this mean for advisors? I hadn't thought about that much, but let me try to provide a quick answer. Perhaps one area is that advisors can focus on other issues (such as building a floor, dealing with some of the emotional aspects of the retirement decision, etc.) with their clients and not get overly worried about choosing a perfect asset allocation.
You mentioned to previous articles that you wrote. Can we provide them
Thanks for that question. I do maintain my Retirement Researcher blog where I will surely provide a link to anything I write in the future. Please consider an email subscription to it (it's free). I am also a monthly columnist at Advisor Perspectives. I put a lot of effort into those columns and sometimes they are mini-research articles. The archives can be found here. And as for research articles, I have them all listed at my university webpage. Most of my retirement planning research is near the top of the list.
Wade Pfau, Ph.D., CFA, is an associate professor of economics at the National Graduate Institute for Policy Studies (GRIPS) in Tokyo, Japan, and the curriculum director for the Retirement Management AnalystSM designation program. He maintains a blog about retirement planning research at
This Website Is For Financial Professionals Only