Linking Academic and Planner Research on Retirement Income

Author: Wade Pfau

Guest Contributor

Wade D. Pfau, Ph.D., CFA, is the Professor of Retirement Income in the Ph.D. in Financial and Retirement Planning program, Co-Director of the American College Center for Retirement Income, and RICP® program director at The American College of Financial Services.

Pfau is a co-editor of the Journal of Personal Finance. He has spoken at national conferences of organizations for financial professionals such as the CFA Institute, FPA, NAPFA, AICPA-PFP, and AFS. He also publishes frequently in a wide variety of academic and practitioner research journals. He is also an active blogger on retirement research and maintains the educational Retirement Researcher website, and you can follow Wade on Twitter.

Wade can be reached by email at: [email protected].

Though I am generalizing somewhat, the planner’s view on retirement income most commonly focuses on safe and sustainable spending rates to draw down one’s financial wealth during retirement. Since the publication of William Bengen’s 1994 article using historical data to provide guidance about sustainable withdrawal rates, discussions have centered on the 4% rule of thumb for sustainable spending. Though vigilance is expected, with a broadly diversified portfolio of stocks and bonds (Bengen suggested 50-75% stocks), retirees should be able to rely on spending an inflation-adjusted amount over 30 years equal to 4% or 4.5% of their retirement date assets. Extending the research to Monte Carlo simulations, probabilistic thinking is encouraged and the focus turns to keeping failure rates sufficiently low.

But the notion of a safe withdrawal rate from a portfolio of volatile assets is somewhat perplexing for academics trained in the theory of life-cycle finance. Their idea is that failure, at least for meeting basic needs, should be eliminated. Life-cycle finance, which Paula Hogan summarized well for JFP readers in 2007, views hedging (such as with a bond ladder) and insuring (such as with fixed income annuities) as important risk management tools beyond precautionary savings (saving enough to use a ‘safe’ withdrawal rate) and portfolio diversification. Life-cycle finance also informs us that current market conditions are much more relevant than historical averages.

Actually, the 4% rule does stand on a rather flimsy basis. It worked during an exceptional and perhaps anomalous period in world history when the United States rose to become the world’s leading superpower. Testing the 4% rule with data for other developed market countries leads to some rather shocking results. Alternatively, especially around the year 2000, market valuations rose to unprecedented levels and we do not have a basis in U.S. history for knowing about sustainable spending from such a starting point. Today the issue centers more on low bond yields, suggesting that we should plan for more modest stock and bond returns than the U.S. averages since 1926.

Markets tend to surprise us frequently, and retirees only get one whack at the cat. With Modern Retirement Theory, Jason Branning, CFP®, and Dr. M. Ray Grubbs inform us that basic expenses should be covered by income sources that are “secure, stable, and sustainable.” For basic needs, retirement income strategies must work by construction. According to the Retirement Income Industry Association (where I serve as curriculum director for the Retirement Management Analyst designation), the fundamental goal of retirement planning is to “first build a floor, then expose to upside.” This is life-cycle finance theory as used in practice.

That being said, I do not suggest abandoning research on safe withdrawal rates. I think we can instead build a more complete framework. Consider the chart to the right. Academics are puzzled about why individuals are not more willing to annuitize their assets. That would eliminate longevity risk. But it is clear that many people are loath to give up control over their assets. That is why inflation-adjusted single-premium immediate annuities (SPIAs) are at the bottom right of the chart. Nominal SPIAs are to their left, since their real spending power may decline dramatically.

On the other hand, systematic withdrawals, from a portfolio focused either on total returns or on a time-segmented bucket approach, are a time-honored and understandable topic for planners and their clients. I rate them as low in terms of protection, but high in terms of control over assets. I should note however, that Bob Seawright rightfully pointed out that this control over assets may be an illusion. Poor market returns, high inflation, and a long life may severely constrain one’s control.

Where research about retirement income strategies is ultimately heading is to the box which includes strategies with partial annuitization, the use of deferred income annuities (DIAs), and variable annuities with guaranteed lifetime withdrawal benefit riders (GLWBs). Moshe Milevsky, Mark Warshawsky, Joseph Tomlinson, and Peng Chen are a few names who are already making an impact in this area. Ultimately, though, the best solutions must incorporate both academic rigor, and the practitioner and client challenges of the real world.

Here are a few ideas about how to better integrate the approaches used by financial planners and academics with regard to retirement income strategies:

• It is time to abandon the idea that retirees should focus on finding a spending strategy that maintains a rather low failure rate, as is par for the course with safe withdrawal rate studies. This narrow focus ignores the lost potential enjoyment from spending more, it ignores how much flexibility retirees may have to cut their spending at a later date, it ignores the availability of other spending resources outside of the financial portfolio (such as Social Security) that will continue even in the event of portfolio depletion, it ignores the low probability of still being alive many years after retiring, it ignores any goals to leave a bequest, and it ignores potentially how long the "failure" condition may last. Historical failure rates also don’t help us if current market conditions are different from the historical averages, and they have no meaning when used with variable withdrawal strategies that reduce spending in order to preserve wealth. We need a more complete model that incorporates these considerations.
• Research must increasingly look at variable withdrawal strategies responding to market returns, as well as how to allocate to bond ladders, SPIAs and other retirement products. For strategies with variable spending, we can investigate the degree of lifetime underfunding from a minimum acceptable or desired spending level. Or even better, we can use measures which incorporate the value of spending (as also discussed by Joseph Tomlinson) and recognize that greater spending provides more enjoyment but at a diminishing rate.
• Academic investigations show that a very powerful rule for systematic withdrawals, after building a floor, is to use a withdrawal rate along the lines of 1 / remaining life expectancy (using the life expectancies provided by the IRS for the Required Minimum Withdrawals from IRAs is fine) for each year of retirement. This can be modified to incorporate bequest motives, a greater desire for consumption smoothing, and a return on underlying assets other than zero. We do also need to pay attention to the uncertainty of potentially large health-related expenses later in life.

It is an exciting time to conduct retirement research closely linked to the needs of planners and clients. I do believe we are at a stage in which we can broaden the safe withdrawal rate question and build a more complete framework for developing retirement income strategies which combine insights from both the academic and planning communities. As Michael always asks, I would love to hear your views.