The practice of systematically withdrawing an inflation-adjusted four percent per year from savings is still the default mechanism that most affluent investors and their advisors use to convert retirement savings to retirement income.
The so-called SWiP method has earned its popularity legitimately. First, it’s simple. In theory—if not practice—you set it and forget it. Second, research shows that a 4% payout from a 60% stock, 40% bond portfolio has a 96% chance of lasting 30 years. Third, it maintains 100% liquidity.
A fourth reason, one rooted in psychology, may also play a role. Average investors probably figure—justifiably or not—that if they withdraw only 4% from a balanced portfolio with an 8% historical return, they’ll achieve the ideal of never dipping into principal.
The 4% approach doesn’t necessarily work in the real world, however. It ignores the fact that some clients retire at age 50 and others at age 70, for one thing. To mend the flaws, imaginative advisers like Larry Frank Sr. of Rocklin, Calif., a small town in the arid foothills of the Sierra Nevada mountains northeast of Sacramento, have developed variations on the classic or “first generation” SWiP method .
Frank uses “a dynamic and adaptive approach to distribution planning and monitoring,” which he described in an article by the same name in the April 2009 issue of the Journal of Financial Planning, co-authored with David M. Blanchett. He advocates revisiting the payout rate every year to adjust for such variables as market performance and the inevitable passage of time.
To create a retirement income plan, Frank meets with the client and establishes a few essential factors, such as likely length of retirement. That can range from 20 to 40 years, depending on age of retirement, health, and mortality tables. If the client wants to retire at age 55, Frank might assume a 40-year retirement and suggest a starting withdrawal rate of only 3%. If the client wants to retire at age 70, he might suggest a 5% starting rate.
Next, he determines how much annual income the client will need in addition to Social Security and pensions. “For instant, if someone needs $40,000 a year in retirement, and they get $20,000 from Social Security, the unfunded portion is $20,000,” he said. Then he divides the required annual income by the starting withdrawal rate to arrive at the target savings amount.
Thus, clients with starting withdrawal rates of 3%, 4% or 5% would need either $600,000, $500,000 or $400,000 at retirement, respectively, to generate an income of $20,000 per year. If they don’t have the magic number by the time they want to retire, Frank said, they need to make adjustments—perhaps by deciding to retire later or live on less.
That’s just the beginning of the process. At intervals within each year of retirement, Frank re-calculates the probability of failure of each client’s account, based on recent performance and the client’s adjusted life expectancy. If the new probability of failure is less than five percent, according to Monte Carlo simulations, he increases the payout rate by 30 basis points (in addition to a standard annual 30 basis-point cost-of-living increase).
Conversely, he will decrease the payout rate by 30 basis points if the probability of failure exceeds certain age-related break points. If the failure risk exceeds 20% and the target end date is 20 years or more away, or if the failure risk is over 10% and the end date is 11 to 19 years away, or if the failure risk is 5% and the end date is 10 or fewer years away, then the payout rate may drop by 30 basis points to reduce the risk of failure. If the client doesn’t want to reduce the payout rate, Frank can reduce the failure risk by tempering the asset allocation.
The two biggest contributions that Frank and Blanchett have made to the original 4% SWiP method, Frank told RIJ, are their annual withdrawal rate adjustments and their decision to use each client’s age of retirement to set their starting withdrawal rate.
The system is somewhat vulnerable to income volatility. “If probability of failure goes down,” Frank said, “those are the years you can take the big vacations. When it goes the other way, then you postpone those things.” On the other hand, he lets clients choose whether they would rather spend less or assume a marginally higher risk of running out of money.
To avoid the trap of reverse dollar-cost averaging, where clients sell fewer shares at higher prices and vice-versa, Frank uses a bucket system. “There’s a long-term bucket that’s for three years and beyond, and there’s a short-term bucket for three years or less,” he said. “I call it their distribution reservoir. [Since the drop in stock values] we turned off the transfers the long-term to the short-term bucket.” To prevent his system from becoming prohibitively labor-intensive, he uses DFA index funds instead of actively managed funds, which relieves him of the chore of monitoring the managers.
Frank started his career as an insurance agent, so it’s not surprising that elements of annuities appear in his version of SWiP. For instance, he slips a kind of mortality credit into his annual adjustments by letting people gradually spend more as they get older. And he acknowledges that his method of establishing a flexible 4% payout echoes the role that an AIR (Assumed Interest Rate) plays in a variable payout annuity.
In certain situations, Frank will suggest that clients consider an income annuity. For instance, if a client’s assets have declined in value and can no longer generate the required income at the suggested withdrawal rate, Frank will tell them that an income annuity could provide the necessary income for less. “An annuity is our backup plan,” he said.
In such situations, Frank has learned some interesting facts about behavioral finance as it applies to annuities.
His clients tend to dread the prospect of approaching the annuity threshold. Indeed, most would rather reduce their standard of living (or keep spending and accept a higher risk of portfolio failure) than buy a life annuity and lose control over a large chunk of their assets. An economist might put it this way: the average person prefers to forego the utility value of the income annuity in order to maintain the illusion that all of his or her assets are truly liquid.
As long as investors embrace that preference, then the SWiP method has a bright future and the “annuity puzzle” will persist.
Mr. Frank, a CFP®, can be reached at [email protected].
© 2009 RIJ Publishing. All rights reserved.
About 60% of MetLife’s variable annuities have GMIB riders. As of last September 30, about 40% of MetLife’s $92.5 billion in domestic variable annuity assets was covered by a living benefit rider. About 70% of the living benefit riders were GMIBs, and 80% of living benefit rider sales in the third quarter 2008 were GMIBs. The most popular version of MetLife’s living benefits is the GMIB “Plus,” which annuitizes for life with a five-year period certain. GLWB and GMAB (guaranteed accumulation benefit) contracts account for the rest.
