Two retirees start with $1 million each, invest it the same way and withdraw a comparable amount every year. One ends up broke. The other has more than $4.5 million. The only difference is the year they retired.
Michael Finke, professor of wealth management at the American College of Financial Services, used that comparison last month to dismantle an assumption he said still shapes too many retirement income planning conversations—the idea that a portfolio’s average return can tell an advisor how much a client can safely spend every year.
A retiree who started drawing down that first million in 2000 watched the account fall to $910,000 within a year and never recovered, Finke said in a webinar hosted by Kitces.com, while a retiree who started the same withdrawals just three years later, in 2003, sidestepped the worst of the early losses entirely. Within these two scenarios, it’s sequence-of-returns risk, not the long-run average, that decides which retiree a client becomes.
That risk is why Finke believes that essential, inflexible spending—the property taxes, utilities, healthcare and food that make up roughly 65% of a higher-income retiree’s budget—should be supported by a foundation of fixed-income investments rather than stocks. More discretionary spending then can carry the real investment risk that also brings higher returns.
Finke laid out three strategies for building that fixed-income foundation, each with its pros and cons: withdrawing from bond funds, building a Treasury bond ladder and transferring longevity risk through an annuity.
Bond funds are the most common approach, he said, and they carry real advantages over stocks, including lower volatility and lower correlation with equities, and sometimes a modest credit premium. But for clients who are looking for predictable income generation, there’s a major drawback: It’s impossible to say in advance how much a bond fund needs to hold to reliably generate a given income.
Finke found those bond funds suffered from the same sequence risks that equities did when he modeled retirees who withdrew $24,000 a year from a $400,000 starting balance in the Bloomberg U.S. Aggregate Bond Index. A 1994 retiree still had about $175,000 left after 30 years, while retirees who started just a year or two later were nearly out of money by the same point, all using the same withdrawal rate.
A Treasury ladder removes that uncertainty by matching specific bonds to specific future spending years, he said, so the total cost is known up-front and isn’t left to the market. Using mortality tables and current rates, Finke calculated that a healthy 65-year-old woman who was guaranteed $24,000 a year until she reached the age of 95 (a point at which she has roughly a one-in-three chance of still being alive), would cost $392,719 today. But that certainty comes at a price.
Locking in today’s rates means locking in today’s regret if rates climb later, he said, and a client watching a 30-year ladder’s market value swing with every rate move may not understand the difference between statement volatility and real income risk. Finke invoked Nobel laureate economist Robert C. Merton, who argued investors chase the wrong kind of safety when they favor short-term holdings like T-bills.
A T-bill’s balance barely moves, but because it must be rolled over again and again at whatever rate happens to be available, the income it produces can swing wildly over a retirement. A ladder does the reverse. Its market value can lurch with rates, yet the income it was built to deliver, locked in at purchase, never changes at all.
Annuities, the third option for supporting retirement spending with fixed income, solve for an unknown that neither bond funds nor ladders fully address on their own—the uncertainty of how long a client will live. By pooling that risk across many buyers, insurers can offer income for a fraction of what a ladder would cost.
Insurers also typically build a credit premium into their payout quotes, Finke said, in effect passing along a return a bit better than Treasurys on top of the longevity guarantee, something he called “almost a bit of a free lunch.” A healthy 65-year-old man with roughly a one-in-10 chance of living to 97 would need about $4,000 today in a ladder to guarantee $15,000 at that age, Finke said, but pooling that bet with nine similar peers through an annuity cuts the cost to about $400 today.
For clients with tax-deferred retirement accounts, he pointed to the additional benefit of sidestepping required minimum distributions on up to $210,000 if it’s moved into a deferred annuity. The downsides of annuities are familiar—a smaller bequest if the client dies early, limited liquidity, and counterparty risk, though state guaranty associations cover annuities up to $250,000 in most states.
Finke said advisors often resist annuities because they believe they can replicate the benefits with investments alone through a total return approach. But he argued that transferring longevity risk to an insurer lets a client fund essential spending with a smaller share of the portfolio, freeing the rest to be invested more aggressively.
“In my own retirement plan, I am going to be doing some sort of blend of investments in annuities, because I understand theoretically the benefit of annuities,” he said.