A new research paper from two well-known investment experts concludes that only people who live much longer than average are likely to benefit from owning an immediate annuity, and that an inflation-adjusted annuity works better than a fixed annuity.
The article also defies conventional wisdom by suggesting that an investor’s allocation to equities should rise in retirement, not fall.
One of the authors, Michael Kitces, has never been a fan of annuities—his high net worth audience can afford to self-insure against longevity risk. But his co-author, Wade Pfau of the American College, has built a reputation on research showing the advantage of income annuities for retirees—even for those who wanted to maximize their bequests. For Pfau, the paper suggests an abrupt change of heart.
RIJ asked Pfau if the paper was a repudiation of his former support for income annuities. In an email, he wrote, “I wouldn’t go that far. It is going to be easy to misinterpret this one, because it was specifically responding to studies which showed SPIAs in too optimistic of a light. SPIAs still provide unparalleled longevity protection and still have a role. Also, we are revising now by exploring more with regard to the magnitude of failure, which is also going to help SPIAs. We will need to revise the article a bit in that regard.”
In their new paper, the two analysts test the wisdom of putting half your money in an single-premium immediate annuity at retirement and the other half in stocks. Their calculations show that many people—especially if they don’t live longer than average—would be better off starting out with half their money in stocks and half in bonds, and then gradually reducing their bond allocation over time.
The authors agree, however, that people who live a very long time will benefit from buying a SPIA with half their money at retirement, and that those who buy an inflation-adjusted SPIA will benefit the most.
“For those who materially outlive life expectancy, SPIAs do continue to show a valuable benefit to improving retirement income sustainability and real SPIAs fare better in such a ‘longevity hedge’ role than nominal SPIAs. However, the required time horizon for real SPIAs to make a meaningful contribution is significant—long enough that only a small percentage of retirees are likely to reach the benefit point, and the crossover takes even longer to achieve given today’s low return environment,” the paper said.
“Most prior studies which indicated a benefit of partially annuitizing a retirees portfolio were actually showing the benefits of a bucketed liquidation strategy that spends down fixed assets first and allows the household equity allocation to rise, not a benefit of the SPIA itself, especially in scenarios that do not extend materially beyond life expectancy.”
The paper recommends higher equity allocations in retirement, not lower. “Not only are declining equity glidepaths potentially harmful, but that surprisingly rising equity glidepaths are actually beneficial,” the authors write.
“The primary scenarios where SPIAs should be used are specifically those were the intent is to hedge significant longevity beyond life expectancy, where SPIAs and their mortality credits simply provide an unparalleled fixed-income-equivalent return. In the remaining scenarios for most retirees, though, the more effective way to improve retirement outcomes is simply to implement a rising equity glidepath.”
Some in the annuity world may find this paper provocative, and not in a good way. Annuities can be viewed through an “investment” frame or an “insurance” frame. This paper seems to evaluate them mainly as investments. Annuities never do make sense as investments; their value shows up mainly as a hedge against extreme aging. Hedges are expensive, but insuring against longevity risk is thought to be cheaper than hoarding against it.
The paper focuses on life-only annuities. That may be a bit of a straw man. Many people who buy annuities buy life-with-period-certain annuities that are guaranteed to pay out most or all of the principal to somebody, if not the original owner. It would be interesting to see what would happen if the authors considered life-with-period-certain contracts.
There’s a lot of implicit faith in equities here; people who don’t believe that stocks always pay off in the long run may question the authors’ belief in the wisdom of rising equity allocations in retirement. The paper also doesn’t appear to acknowledge that the severe pain of running out of money in extreme old age, in addition to the probability, adds weight to the value of life annuities.
© 2013 RIJ Publishing LLC. All rights reserved.