Retirement Planning Assumptions Revisited

This analyst blog is part of our coverage of the 2016 Morningstar Investment Conference.

During the final breakout session of the 2016 Morningstar Investment Conference, Christine Benz moderated a discussion with Jonathan Guyton of Cornerstone Wealth Advisors and David Blanchett, who heads up retirement research for Morningstar asset management, about their respective research.

Benz started by asking Blanchett about the 4% withdrawal rate for a retirement portfolio. He provided some context by explaining how the figure came from the 1994 paper by Bill Bengen, which showed that over a 30-year period, a starting withdrawal amount based on 4% of the portfolio (and adjusted for inflation) was a safe starting point for most asset allocations. He provided a chart that revealed, with the exception of a portfolio with less than 20% equities, the probability of success was approximately 90% for all asset allocations.

Blanchett clarified that the guideline didn't look at ending balance, as the Bengen study only addressed the safety of funding retirement for a 30-year period. He also noted that the 4% meant the initial withdrawal amount, not 4% of the portfolio each year. However, these success rates are all based on historical return rates. When using projected future returns, which aren't as rosy, prospects are much dimmer. A balanced portfolio has approximately a 50% chance of success, a "coin flip," as Blanchett put it, while an all-bond portfolio has only a 3% chance of 30-year success. He suggested that we've overstated the safe withdrawal rate because the U.S. market has done so well for investors in recent years.

He also reminded the planners in the audience that their clients are generally wealthier than the average person, and hence likely to live longer, resulting in a longer retirement period. Benz asked Blanchett what he considered an acceptable rate of success and he said it doesn't make sense to play it too safe, such as planning for a 40-year retirement with a 99% success rate, as it means clients will likely die with many of their assets. He said he generally likes plans with an 80% chance of success.

When asked about expected returns, Blanchett said he expects a nominal return of 6.5% for large-cap stocks, which would then be diminished by inflation and fees. Guyton concurred with Blanchett's expectation of lower future returns, but noted that some clients might actually be retiring in better-than-expected circumstances, as recent market performance might well have boosted portfolio balances beyond what a planner might have projected five years ago.

Guyton then addressed research he and his colleagues have done about adjusting equity exposure of portfolios based on current valuations, with the idea of reducing equity stakes if valuations are high. He said their current recommendations reduced equity exposure by about one tenth, compared with periods of normal valuation. He also said it was worth asking to what degree does getting this assumption right matter, and what are the consequences if you give advice and it's wrong. He noted that there are three approaches one can take--basing advice on what worked in the past; future-looking, which requires a prediction; or a third approach, where instead of past performance or future predictions, he draws on signals the market and clients' spending are giving him.

Benz's question about expected inflation rates sparked a discussion about CPI-W versus CPI-E, with Blanchett noting that the inflation rate really depends on individuals and their specific purchases. The conversation then turned to retiree spending, and the fact that it's not constant throughout retirement. He illustrated this trend with a chart showing that for different income levels, consumption in retirement drops throughout retirement before an uptick at the end of life.

Guyton said accounting for this spending decline allows him to take a different approach with his retired clients. That is, by setting aside a portion of the retirement portfolio sufficient to fund the end-of-life spending amount, in excess of Social Security, it allows him to designate another pot of money as fully discretionary, to be spent as the client sees fit. This approach gives more flexibility to the client, and means the advisor doesn't have to be the "bad cop" or the ongoing decision-maker.

Benz raised the issue of dynamic withdrawal during retirement, and Guyton noted his clients "do not spend money as if they are a spreadsheet." The discussion then turned to research he published in 2006 with William Klinger that used a "guardrail concept" to establish decision rules by which retirees could increase or decrease portfolio withdrawals based on market conditions and portfolio performance.