Timely, relevant, and actionable investment perspective, best practices, and planning insights for institutional and wealth management clients from CAPTRUST's Consulting Research Group.

Advanced Filter

Mastering Retirement Decumulation

John Curry
Senior Director | CAPTRUST Marketing

Scott Matheson
Senior Director | CAPTRUST Defined Contribution Practice Leader

We have often suggested that the best way to help prepare employees for retirement is to get them in a retirement plan, get them saving enough, and get them invested well. Plan auto features combined with well-constructed default investment options will go a long way to help many participants, especially younger ones with time on their side, accumulate meaningful wealth for retirement. However, when it comes to retirement income generation — actually turning on the spigot when they cease to have income from work — many may not be so well prepared. And with roughly 10,000 baby boomers retiring every day for the next 16 years,1 the “how” of retirement income generation is not a trivial challenge. While successful retirement income generation can be nuanced and doesn’t lend itself to simple answers, the solution lies at the intersection of income replacement math, capital market expectations, and longevity.

We generally recommend that retirement savers accumulate enough wealth to replace at least 70 percent of their pre-retirement income as they enter retirement. To put this into perspective, an individual earning $50,000 a year should be in a position to replace at least $35,000 a year to be considered ready to retire without sacrificing lifestyle. If she assumes that $15,000, or approximately 30 percent of her retirement cash flow will come from Social Security, she will need to generate an additional $20,000 a year from her retirement portfolio. 

Thirty years ago, she could have entered retirement invested in a safe, all-bond portfolio with long-term yields approaching 10 percent. Such a portfolio would have provided $10,000 per year for every $100,000 of savings. She would have needed approximately $200,000 in the all-bond portfolio to fund a 30-year retirement. Interest rates at those levels meant that retirees could live off of yield alone, ostensibly keeping their principal intact — albeit exposed to inflation’s erosive effect and without a lot of upside potential for overall portfolio gains.

Since then, we have witnessed a prolonged bull market for bonds, as the yield on the 30-year Treasury bond has fallen from a high of 15.21 percent in October 1981 to a low of 2.45 percent in July 2012. The frequently quoted 10-year Treasury has followed suit, with its yield falling from 15.84 percent to 1.39 percent over a similar time period. Needless to say, the days of 10 percent yields are long gone, and most investors can no longer depend on yield alone to fund retirement spending.

Today’s low-interest-rate environment requires a different approach to retirement investing and income generation. In fact, “income generation” is really a misnomer. It’s actually more accurate to call it withdrawal management. This concept is also sometimes called decumulation, a recently coined term meaning the opposite of accumulation. Withdrawal management is defined as the orderly liquidation of retirement assets, generally a globally diversified portfolio combining stocks and bonds, over the course of one’s retirement. Generally speaking, withdrawal management entails taking a formulaic approach to annual distributions from or liquidation of one’s retirement portfolio. There are many schools of thought on the various formulae: which strategy is safest, which produces the most retirement cash flow, or which best models the reality of retirement spending. In the end, as with many things in life, what is “best” lies very much in the eye of the beholder.

Given CAPTRUST’s long-term capital market expectations and a portfolio consisting of 40 percent market-driven or directional assets and 60 percent non-directional assets, withdrawal rates between 3 percent for a 30-year period of retirement and 4 percent for a 20-year period are reasonable according to our research. Put another way, every $100,000 of retirement savings should be able to reliably provide $3,000 per year of inflation-adjusted cash flow for 30 years of retirement or $4,000 per year for 20 years. 

Figure One shows the sustainable withdrawal amount associated with a range of periods, from 10 to 40 years. As you can see, a longer retirement period requires relatively low withdrawal amounts, while shorter retirement periods tolerate higher withdrawal levels.

The data in Figure One suggests that our hypothetical $50,000-a-year employee will need a portfolio between $500,000 ($20,000 divided by 0.04) and $667,000 ($20,000 divided by 0.03) to safely fund her income replacement, assuming she is planning for a retirement period that lasts between 20 and 30 years. That’s two and a half to three times more retirement savings needed to fund the same income level than in the previous all-bond portfolio example.

Funding retirement via withdrawals from a total return portfolio is by necessity fundamentally different and riskier than simply harvesting yield from a bond portfolio. However, this approach can help to maximize the amount of retirement cash flow a retiree can generate, given today’s low-interest-rate environment. Nonetheless, a few caveats:

1. This analysis is based upon a series of simulations using forward-looking capital market return and risk assumptions, which could prove to be incorrect. In the words of baseball player and philosopher Yogi Berra, “It’s tough to make predictions, especially about the future.”2

2. The simulations performed in our research provide an understanding of the range of possible outcomes for a variety of time periods. For each period, we modeled 1,000 potential future outcomes. Some of the potential outcomes turn out well, with investment returns outpacing withdrawal needs. Others turn out not so well, failing to provide adequate cash flow for the complete time period. Our reliability measurement for a withdrawal strategy was a 90 percent likelihood of success generating inflation-adjusted withdrawals for the analyzed time period. In other words, we targeted withdrawal strategies that worked in 900 or more of the 1,000 simulations we ran.

3. A retiree employing this type of strategy may invade or, in some cases, grow principal depending on how the capital markets behave. Growing principal too much may be looked at as a lost opportunity for retirement withdrawals or spending. Invading principal may be a lost opportunity to create a legacy for kids, grandkids, or a favorite charity.

Longevity, a topic detailed in last quarter’s Strategic Research Report, is another issue to consider in retirement withdrawal management. The decline in availability of defined benefit pension plans means the elimination of widely available guaranteed retirement income (beyond that provided by Social Security). This forces individuals to plan for the probabilistic aspect of longevity by anticipating a longer-than-average lifespan and period of retirement. The life expectancy of a 65-year-old female in 1980 was 18.3 years; 14.1 years for a male. By 2010, those numbers had risen to 20.3 and 17.7,3 respectively, meaning that individuals are now tasked with funding two or three additional years of retirement. To further compound the problem, life expectancy is expected to continue to increase in developed countries like the United States at a rate of 2.5 years per decade.4

The remedy to these concerns is recognizing that withdrawal management is not a set-it-and-forget-it endeavor. This is in distinct contrast to the accumulation phase, when putting a plan in place and sticking to it can be a primary determinant of success. But when it comes to decumulation, as the saying goes, “no plan survives contact with the enemy.”5 Any strategy put in place should be subject to change based upon market conditions, investment results, and changes in personal circumstances. For example, a few good years of investment results may create a larger portfolio that — all things being equal — could support a higher level of withdrawals. Keeping the 

The simulations performed in our research provide an understanding of the range of possible outcomes for a variety withdrawal level the same on this larger portfolio will provide more outcome certainty and perhaps a larger legacy for beneficiaries or will allow the investor to reduce overall portfolio risk.

By the same token, spending patterns change over time, and withdrawals should be modified to address these changes. For example, the initial retirement years may include considerable travel and entertainment expenses—reflecting, perhaps, a newly retired couple’s desire to take advantage of good health and financial wherewithal, following pursuits they may have neglected during their period of career work. During this early retirement period, they may exceed the 3–4 percent “speed limit” of a prudent withdrawal strategy, knowing that they are highly likely to slow down spending after this initial period of making up for lost time. And, of course, healthcare-related events throughout the course of retirement may negatively affect both lifestyle and household finances and may require further withdrawal adjustment.

The Pension Protection Act helped legitimize auto-enrollment, auto-escalation, and qualified default investment alternatives beyond capital preservation options—features that are helping plan sponsors get employees in a plan, saving enough, and invested well. Savings accrued over years of retirement plan participation, on top of a foundation of income expected from Social Security, may make retirement an attainable goal for many. However, it may not be sufficient to leave them successfully at the point of retirement. Retirement decumulation presents an additional challenge participants and plan sponsors should be aware of. The combination of capital market expectations, continued low current interest rates in particular, and increasing longevity will require participants to take a thoughtful and fluid approach to withdrawals to fund spending during retirement. We welcome the opportunity to speak with you more about this important topic. 


1 Pew Research Center. “Baby Boomers Retire.” Pew Research Center. n.d. Accessed December 2013
2 The Economist. “Letters to the Editor — The Perils of Prediction.” The Economist. 2007. Accessed December 2013
3 National Center for Health Statistics. “Life Expectancy at Birth, at Age 65, and at Age 75, by Sex, Race, and Hispanic Origin.” Centers for Disease Control and Prevention. 2012. Accessed December 2013
4 Max Planck Institute for Demographic Research. “Survival and Longevity.” Max Planck Institute for Demographic Research. n.d. Accessed December 2013
5 Hartford, Tim. Adapt: Why Success Always Starts with Failure. New York, NY: Farrar, Straus and Giroux, 2011