A new, award-winning research paper warns advisors that the financial planning models they are using fail to match retirement reality.
The more sophisticated retirement-oriented planners are aware of “sequence of returns risk” and incorporate gyrating return possibilities into their clients’ portfolios.
But, argues Dirk Cotton, author of the new paper, advisors are likely to be understating retirement income volatility by not factoring in what he calls “sequence of consumption” risk.
In late May, the Retirement Income Industry Association (RIIA) conferred its 2015 RIIA Thought Leadership Award on Cotton for his paper published in the spring issue of RIIA’s peer-reviewed Retirement Management Journal.
While Cotton, the principal of Chapel Hill, North Carolina-based JDC Planning and author of a retirement wonk-oriented blog called The Retirement Café, tells ThinkAdvisor that advisors should add sequence of consumption to their retirement risk modeling, he first cautions that sequence of returns is a misleading term.
Usually used to describe the risk of portfolio shortfall based on unpredictable market timing — in a case where, for example, the market crashes the day your client retires — Cotton argues that advisors would do better to divide sequence of returns risk into two conceptually more precise terms: “path dependency” and “probability of ruin.”
The former term simply means that the volatile assets your client owns will go up or down unpredictably. So path dependency may be the reason your client’s retirement portfolio shoots up rather than down if the retirement occurs at the start of a long bull run, for example.
Or, a retiree who doesn’t have path dependency may run out of money before running out of time.
As an illustration of the latter case, Cotton gives the example of someone who is blessed with an especially long life of, say, 105 years.
“Did they run out of savings because of path dependency? No. The portfolio lasted as long as it could.
“Would you blame the portfolio failure on sequence of returns risk?” he asks. “I don’t think so. The probability of ruin is because of longevity.”
That is why sequence of consumption is so important, yet the existing financial models don’t address this reality.
The most popular model relied on by financial advisors calls for systematic withdrawals, the most famous expression of which is the controversial “4% rule,” which sets that figure as the safe annual withdrawal rate.
Although an advisor might perform an analysis showing that, say, 95% of the time, a client can draw down funds at that rate and not run out of money, “I would suggest and common sense dictates that most people don’t spend like that. That is not a very realistic model,” Cotton says.
The retirement researcher — Cotton says he himself has been retired for 10 years and only sees a limited number of clients “for enjoyment” — said his own research shows that clients would obtain better outcomes were they to take a fixed percentage not of their starter portfolio but of their annually recalculated portfolio balance.
“It goes up and down every year, but you almost never run out of money,” he says.
Yet that method too is not realistic for most retirees.
“What is common,” he says matter-of-factly, “is you spend the amount of money you need to spend, not the amount you calculated at end of last year. If your daughter has appendicitis, you don’t say ‘Sorry dear, we’ll get that operation next year.’”
Similarly, if the portfolio generates returns above expectations, he continues, “you don’t tell your wife, “We’re going to the Caribbean this year.”
The reality that retirees spend what they need to spend, readjusting down the road if they must, signifies that retirement plan modeling must incorporate a second crucial variable.
Not only must advisors plan for sequence-of-returns risk — or better, path dependency — recognizing that a gyrating market makes portfolio values unpredictable; so too must they plan for sequence of consumption risk.
“Spending is [also] a stochastic process — it doesn’t result in a predictable value … So sequence of consumption risk means really acknowledging that not only the returns on the portfolio are random and unpredictable, but your spending is random and unpredictable.”
Again, sequence of consumption is not necessarily “bad” any more than sequence of returns is. A person may spend less than what was originally planned and thus have a higher than expected portfolio value.
The key insight is that advisors must grapple with double the amount of randomness from both gyrating portfolio values and unpredictable and unpredicted spending patterns.
So what’s an advisor to do?
Cotton recommends two things: First, get the model right.
“You get better planning results if you use a more realistic model.” The model suggested by the 4% rule is “not true,” he says, because it assumes you will be spending a flat amount of money for the rest of your life.
But that won’t happen if you go broke 10 years into a 30-year retirement plan; some clients just have to have that Cadillac Escalade and second home, or a loved one is stricken with a medical condition requiring costly treatment.
Second, Cotton confronts the double dose of randomness through “dynamic updating.”
Quite simply, advisors and clients should re-evaluate their retirement plan annually based on how much money was actually spent, changes in life expectancy, revisions in future expected spending and the like.
A sailor in his off time, the retirement wonk says the annual update is no different in concept than the daily nautical recalculations you would make if you set sail from Annapolis to Freeport. One naturally drifts off course and must therefore frequently recalculate.
And because of the latent danger of exhausting one’s portfolio, Cotton recommends advisors take a “floor and upside” approach to compensating for the probability of ruin.
“There’s always a chance you’ll go broke; so you have to … have a floor of safe non-market-depending income assets like Social Security, TIPS, bonds or annuities that provide income no matter what the market does,” he says.
— Related on ThinkAdvisor: