The 4% withdrawal strategy, a strategy used to produce income in retirement by many financial advisers and do-it-yourselfers, is, to put it bluntly, oversimplistic. So say the authors of a new white paper from Merrill Lynch's Wealth Management Institute.
"We find this rule to be overly simplistic," said the authors of Merrill's white paper. "Sustainable spending rates depends critically on a (person's) age, gender, and risk tolerance."
Before we get into the weeds, it's worth noting that there are three main ways to produce income in retirement, according to Merrill's white paper and many other experts on the subject: the systematic withdrawal program or the 4% rule, the time-based segmentation approach, and income floor with upside approach.
And each approach has its advantages and disadvantages. But the approach of choice for many financial advisers happens to be the 4% plan, which, in essence, says you can realistically afford to spend 4% of your wealth in retirement each year.
Unfortunately, this approach -- which was first studied in 1994 by William Bengen, and then in 1998 by Philip Cooley, and which is now being scrutinized in countless studies these days -- doesn't deal with all of the risks retirees face in retirement, including and especially longevity risk (the risk of outliving your assets), shortfall and the magnitude of shortfall risk, sequence of return risk (the risk of withdrawing money in down markets), asset allocation risk, and spending risk.
Truth be told, many advisers and do-it-yourselfers don't blindly follow the 4% rule. They make adjustments to rule. But these adjustments are sometimes based more on what some economists call "junk" science rather than real science.
For instance, some advisers use something called a "dynamic" withdrawal strategy where the amount withdrawn is largely a function of market returns. You withdraw less when markets underperform expectations and more when markets outperform expectations.
Unfortunately, this twist on the 4% rule, even with all the hubris on the part of financial advisers who embrace this rule, still doesn't address all the risks of retirement.
Enter Merrill's white paper about the subject, which the authors said explicitly accounted for longevity risk, offered specific asset allocation advice based on a person's age and gender, and considered not just the likelihood of a shortfall, but also the potential magnitude. And the bottom line, said the authors, is this: "There is no one-size-fits-all guidance for spending rates."
According to the authors of Merrill's white paper, there are many factors that affect retirement outcomes. But only two of those factors fall under a person's complete control: spending levels and asset allocation.
And of all the factors, the one most crucial to assuring retirement security is spending. "Just as saving aggressively is the key to accumulating wealth, not overspending is the key to safeguarding retirement," the authors of Merrill's white paper wrote. "Asset allocation also matters, but far less. Retirees therefore need to know what spending levels are prudent so that they can enjoy their wealth over their lifetimes -- but not too much too soon."
So, at what rate can a retiree sustainably spend? And what asset allocation best facilitates this spending? The answer, said Merrill's experts, depends on a person's age and asset allocation.
In its study, for instance, Merrill ran one set of simulations assuming a 65-year-old woman had a $1 million portfolio with 60% in stocks and 40% in bonds. With an initial spending rate of 2.5% or 3%, the authors said it's unlikely that the woman would run out of money over her lifetime.
In fact, odds were high that this woman would be able to leave a bequest of $1 million or more. But if her initial spending rate was say 5%, she'd have a one in four chance of outliving her portfolio, according to Merrill's analysis. "Excessive retirement spending may be a hazardous to a (person's) wealth," wrote the authors of Merrill's white paper.
The other factor within a retiree's control, according to Merrill, is asset allocation. In Merrill's example, if the 65-year-old woman had a spending rate of 4% and an all-bond portfolio with no money in stocks, her worst-case shortfall would be $360,000.
If, however, the woman allocated 40% to stocks and 60% to bonds, her worst-case shortfall would be $170,000. And raising the percent allocated to stocks above 40% increased the worst-case shortfall. In other words, in Merrill's example at least, allocating anywhere from 0% to 40% represented the ideal asset allocation.
That range, the authors of Merrill's white paper wrote, represents "a 'sweet spot' over which increasing the allocation to equities boosts the median bequest while mitigating the risk of a worst-case scenario."
As long as the (person) is comfortable with the added volatility of stocks, the woman in this example stands to gain by having some equity exposure. "Allocating some portion of a retirement portfolio to stocks can increase a (person's) potential bequest while reducing her risk of outliving her wealth," said the authors of Merrill's white paper.
By the way, the authors of Merrill's white paper used the following expected returns in making its asset allocation recommendations: U.S. stocks, 9.5%; U.S. bond, 5.3; and U.S. cash, 3%. And it used mortality assumptions from the Society of Actuaries, Annuity 2000 Mortality Table.
Of course, one of the factors (spending or asset allocation) in each of the above examples is fixed in each example.
And that's not how things work in reality. So, one has to examine the interplay between spending and asset allocation and the probability of your money lasting a lifetime, according to Merrill. And when viewed through that lens, the authors of Merrill's white paper suggested that a woman age 65 with a $1million portfolio had a 95% chance of not outliving her money if she had an initial spending rate of 3.5% and allocated roughly 20% to 70% of her assets to stocks.
Your personal "achievable spending rate" will, of course, be based on your age, gender, risk tolerance, and how much certainty you desire, according to Merrill's white paper.
For instance, a 65-year-old woman who wants a moderate certainty (a 90% probability of success) of not outliving her assets might spend 3.9% of her wealth. By contrast, if she wanted a high certainty (a 95% probability of success) of not outliving her wealth might spend just 3.5% of her wealth.
By way of background, there is much debate over how much retirees should invest in stocks, with many suggesting very little and others suggesting quite a bit. For instance, Putnam Investments' research group recently suggested that retirees allocate just 5% to 25% of their portfolio in stocks if they want to minimize the risk of running out of money and sustaining their withdrawals.
Read Robert Powell's column from July, 2011, Retirees need fewer stocks, more annuities.
So what do those who are currently studying the 4% rule have to say in reaction to Merrill's white paper? Not surprisingly, academics and planners had differing points of view.
"This article is a pretty standard safe withdrawal rate analysis and it doesn't do anything new," said Wade Pfau, an associate professor at the National Graduate Institute for Policy Studies and co-author of several studies on the systematic withdrawal plan approach.
Among Pfau's quibbles is Merrill's use of mortality data vs. planning for a long horizon. Pfau has examined both approaches and said he prefers using the latter approach. "With mortality data, one way to get success is to die quickly, which seems like an unhelpful way to plan for retirement," said Pfau, who just co-authored a paper on the subject.
(Read his paper, Spending Flexibility and Safe Withdrawal Rates, which was just published in the Journal of Financial Planning, at this website.)
"But all in all, this study is fine," said Pfau. "It isn't unique, as it does show the normal conclusions from safe withdrawal rate research."
For his part, Joseph Tomlinson, a principal of Tomlinson Financial Planning and author of "A Utility-Based Approach to Evaluating Investment Strategies," a recently published paper in the Journal of Financial Planning, is among those who favor allocating a higher rather than lower percent to stocks in retirement, especially if you have designs on leaving a bequest, an inheritance.
"I happen to think it's important to focus on the tradeoff between bequest motivation and shortfall risk, and those kinds of considerations should be central to this kind of analysis," said Tomlinson. "The approach of prescribing allocations where only the shortfall risk is considered may distort things toward lower-than-optimal equity allocations. This would be particularly true for clients that have a family support system in place where 'mom and dad' have some downside backup and are strongly motivated to leave a sizeable bequest for the kids -- not an unusual situation for investment company clients."
Tomlinson also noted that Merrill made a classic assumption in its paper that retirement expenses increase at the inflation rate. "That may make sense for basic living expenses, but if retirees have some discretion over expenses, a declining pattern may be more appropriate," he said.
A potential drawback to Merrill's white paper is that asset-class return assumptions appear to be based on about a 2% underlying real interest rate. "Currently that rate for U.S. investments is about zero," Tomlins said. "Perhaps they are prescribing a strategy for an investment environment different than the current one -- more in line with long-term historical averages -- but they should have been more clear about that, and at least addressed the discrepancy between their assumptions and the current environment."
All that nitpicking aside, Tomlinson agreed with the general conclusion of Merrill's white paper: "That what may seem safe to some conservative investors -- all bonds or all CDs -- really isn't the safest approach to investing for retirement."
Meanwhile, Duncan Williams, a graduate part-time instructor at Texas Tech University's Division of Personal Financial Planning, agreed with Pfau's assessment that Merrill's paper is not unique. "But given that this is Merrill Lynch paper, I do find it significant that this study is accounting for longevity risk and that the study considers magnitude of shortfall in addition to likelihood," Williams said. "This paper is evidence that big firms are considering making these changes to their safe-withdrawal rate analysis. And if you compare this paper to the equity recommendations in this paper to those in Putnam's paper (link above), you will find that there is a trade-off between likelihood and magnitude of failure. In general, more aggressive approaches, up to a point, are less likely to run out of money, but in the left tail, the magnitude of failure is more severe."
Williams, as did Pfau, took issue with Merrill's use of mortality tables vs. a fixed time horizon. "If one had to choose a fixed period with which to model, any risk-averse person would handle longevity risk by adopting a horizon that is longer than their life expectancy," said Williams. "So the results may not be that different than a utility-based model that models life expectancy using the mortality data. Using a long fixed period model artificially offsets the effect of not considering longevity risk aversion by overweighting the probability of experiencing decreased consumption at later ages."
Williams said his preference is to include both risk aversion and uncertainty in the model. "I think the improvements to the safe withdrawal approach are about giving a framework for making more informed tradeoffs in the face of uncertainty," he said.
Williams also noted another drawback to Merrill's white paper. "The main issue that I have with stand-alone safe withdrawal rate analysis is that is does not consider other resources and the ability to purchase annuities," he said. "But despite missing some things, I view the industry moving forward in their modeling as a positive sign."