Want to trigger an impassioned debate? Ask a group of advisors about the choice between systematic withdrawal plans (SWPs) and single-premium immediate annuities (SPIAs). Fee-only advisors are loath to cede control of client assets to an insurance company that might someday default, while annuity advocates fire back that only their strategies provide a lifetime income guarantee.
While I’m not taking sides in this feud, I would like to highlight some of the key issues that divide these opposing camps. These include the value of the historical record, the validity of research assumptions, biases in advisor recommendations, the role of fraud and cognitive decline, taxes, and the potential for self-annuitization.
A client following an SWP invests her assets at the start of retirement and withdraws funds over her lifetime. The key determinants of whether such a plan will succeed – meaning that the client will not outlive her funds – are the rate-of-return on the invested assets and the rate at which the client withdraws funds. By contrast, a retiree could instead purchase a SPIA from an insurance company and receive known payments for the rest of his life.
Basic cases
The basic case for systematic withdrawals is that, by calibrating a client’s withdrawal rate to what would have always worked in the past, a client can obtain a sustainable income stream for life similar to what SPIAs currently provide. Even better, systematic withdrawals allow clients to maintain control over and flexibility of their assets, and they retain the ability to bequeath any remaining assets. More often than not, using a safe withdrawal rate will allow a client’s wealth to continue multiplying throughout retirement, except in the unfortunate case when that client’s portfolio suffers a sharp decline shortly after the onset of retirement, which is the absolute worst-case scenario for such investors.
Michael Kitces of the Pinnacle Advisory Group recently emphasized in an important blog entry that safe withdrawals rates also provide a floor-with-upside approach, with the potential for greater spending or a growing legacy. What’s more, aside from the uncertainty of health expenses, which complicate any strategy, available evidence suggests that discretionary expenses decrease with age. With some flexibility to make mid-course spending adjustments, systematic withdrawals from a well-diversified portfolio should safely support a spending plan without wealth depletion while also preserving the client’s liquidity and upside potential.
As for SPIAs, each client only gets one opportunity to enjoy a sustainable retirement, and averages are irrelevant when dealing with volatile investments. With an SWP, one may hope for a risk premium from equities, but there is no guarantee that markets will comply. Especially when considering basic spending needs, relying on withdrawals from a volatile SWP portfolio exposes retirees to greater market risk than they may realize. The U.S. record is much too short to have any confidence about a safe withdrawal rate, uncertainty that is amplified in today’s low interest rate environment. The consequences of being unable to meet basic needs because of a bad sequence of market returns could be severe.
While systematic withdrawals force retirees to plan for a longer lifespan than average, the mortality credits provided by annuities (those who die sooner subsidize the payments to those who live longer) allow for payouts to be connected to life expectancies. Annuities provide a risk management tool that helps to protect clients from sequence-of-returns, longevity and market risk.
So there are countervailing benefits to each approach. The question, I’d argue, is not whether to annuitize, but how much to annuitize. Most Americans will find that the optimal answer is not 0% or 100%, but somewhere in between.
An illustration
Figure 1, below, illustrates the link between real interest rates and annuity payout rates. I adjusted the fees on SPIAs so that payouts are fairly realistic. The figure also includes sustainable withdrawal rates over 30 and 40 years under a simplifying assumption that real portfolio returns will be fixed at the interest rates shown on the axis for the entirety of the retirement period. For SWPs, this is of course unrealistic, as a bad sequence of returns in early retirement could result in lower payouts than implied by a fixed average return. But a diversified portfolio would, on average, support a higher withdrawal rate than what the figure shows as based on current real yields reflecting future portfolio returns.
The figure illustrates three key concepts: (1) mortality credits allow for higher payouts, because the annuity provider can plan for customers to survive to their life expectancy, whereas those planning for systematic withdrawals must assume and save for a longer lifespan to avoid outliving their wealth; (2) sustainable withdrawal rates are closely linked to expected portfolio returns, and simply looking at what worked in the past may not be sufficient, especially as bond yields are currently at historic lows; and (3) current low interest rates do not imply that SPIAs are unfavorable at present, as any strategy must accommodate lower interest rates.
The value of the historical record
For both sides of the debate, important points of contention center on the underlying support provided by the historical record. Annuity supporters argue that the U.S. historical record is much too short to reliably determine sustainable spending rates from portfolios of volatile assets. Even in the wake of such events as the Great Depression and the Great Stagnation of the 1970s, U.S. markets have generally recovered quickly, but, as I have explained in a past column, safe withdrawal rate guidelines based on U.S. data have not performed nearly as well in most other developed-market countries.
For advocates of systematic withdrawals, this criticism of the historical record may be tolerable as long as it is well understood that the critique applies just as much to annuities. Guarantees should not be accepted at face value. Any market scenario adverse enough to jeopardize an SWP would also affect the balance sheets of annuity providers. Credit risk for annuity providers is a real concern, and clients must be aware of potential risks to their guaranteed income. During a systemic crisis, state guarantee associations may be overwhelmed. For some SWP advocates, mistrust of the insurance industry leads to the belief that there are scandals or misdeeds waiting to be uncovered, or that the guarantees built into SPIAs are not as safe as their supporters assert. Joseph Tomlinson provided a recent deeper analysis of annuity-provider credit risk.
Validity of safe withdrawal rate research assumptions
Annuity supporters can also make the point that the basic underlying assumptions for safe withdrawal rate studies do not match the experience of real-world investors. Clients must pay fees to their portfolio managers and investment advisors. As well, the underlying 30-year planning horizon for the 4% rule may not be long enough for today’s retirees. The probability of surviving beyond that timeframe is rising because of medical advances.
Investors must also deal with the tracking risk that their own investment returns may not match those of the precisely rebalanced underlying indices of the hypothetical retiree. For historical safe withdrawal rates to be relevant, clients must rely on very low-cost index funds and rebalance periodically while avoiding temptations to buy high and sell low. The implication is that the hypothetical withdrawal rates found in existing studies are too high and cannot realistically be compared with annuity payout rates. As Bob Seawright of Madison Avenue Securities wrote in The Annuity Imperative, “The reality is that there will be a significant number of catastrophic outcomes using the 4% rule. This alleged safe withdrawal rate is anything but.”
Biases in advisor recommendations
Retirement planning is a process whose goal is to find the best solutions for each client’s personal circumstances. But in the debate between systematic withdrawals and annuities, a central point of contention is how often recommendations are based on the personal interests of advisors more so than what is in the best interests of their clients.
Advisors working for insurance companies are more likely to recommend annuities or other insurance solutions, while fee-based advisors are more likely to suggest systematic withdrawals from a portfolio of assets. Advisors who charge fees for assets under management act against their own self-interest by recommending annuitization, since fewer assets will remain on which to assess fees. SPIAs also limit the frequency of future contacts between clients and advisors, a disincentive for advisors whose fee structures are based on provided services.
This is a tough issue to adjudicate. One certainly hopes we can expect that good advisors will put their clients’ interests first, though research in other areas of finance and psychology has increasingly demonstrated that even unconscious biases can have pervasive effects on financial decisions.
Fraud and cognitive decline
An argument in favor of annuities is that by locking up savings with fixed payouts, a client is better protected from the travails of cognitive decline, which could someday limit his or her ability to make complex decisions about withdrawals and suitable investments, and which might increase the likelihood of falling victim to financial fraud.
Recent research, such as this article by Texas Tech University and University of Missouri professors Michael Finke, John Howe, and Sandra Huston, suggests that cognitive skills for making financial decisions decline with age, while confidence in one’s ability to make those decisions may even increase slightly. Given such findings, Harvard economist David Laibson has referred to annuities as “dementia insurance”. Though fraud could also lead to the siphoning of annuity payments, the process would move more slowly and might prove easier to thwart than the one-time capture of an elderly client’s financial assets.
Another consideration is that when the spouse who took primary care of investments dies, a joint and survivor’s annuity provides a smoother transition for the surviving spouse, who may not be as familiar with investing.
Those opting for an SWP can still mitigate such considerations, of course. Though flexibility with spending is desirable, make sure that a plan is put in place while clients still have full command of their faculties. And there are methods for advisors to continue helping a client with reduced capacities without relying on annuities, such as a medical power of attorney or a waiver allowing an advisor to discuss the client’s finances with others.
Taxes
Taxes are another important consideration, which Joseph Tomlinson delved into extensively in a recent column. To briefly summarize this important issue, SPIAs benefit from deferring taxation by spreading out income over time, whereas systematic withdrawals will tend to frontload the taxation, since the portfolio and accruals are larger early on. However, SPIAs treat all interest as ordinary income, a disadvantage given that some portion of portfolio withdrawals will benefit from lower dividend and long-term capital gains tax rates. The portfolio cost basis at time of retirement is also important; clients must watch for a potentially large tax hit if selling assets to annuitize. These complexities call for an independent judgment for each client about which strategy is more tax-efficient.
Potential for self-annuitization
Advisor Perspectives’ own Robert Huebscher considered safe withdrawal rates from portfolios other than the typical diversified collection of stocks and bonds. He identified municipal bonds as securities that may offer similar cash flows and risk characteristics as an annuity. For retirees who are able to find such bonds with coupons matching the post-tax income provided by the annuity, it makes sense to buy the bonds instead. Payouts would be the same over the maturity of the bonds, and the bonds also provide liquidity and the return of principal.
The bottom line
Given the complexities described above, in this debate I tend to side with noncombatants such as York University finance professor Moshe Milevsky, who advocates retirement portfolio construction that consists of allocating resources among a traditional diversified portfolio, SPIAs, and variable annuities with guarantee riders. Only in a rare case would a retiree’s optimal strategy rely on just one of these components.
Rather than choosing sides, advisors can best serve their clients by becoming familiar with the benefits and disadvantages of each retirement income tool.
Wade Pfau, Ph.D., CFA, is an associate professor of economics at the National Graduate Institute for Policy Studies (GRIPS) in Tokyo, Japan. He maintains a blog about retirement planning research at wpfau.blogspot.com
Read more articles by Wade Pfau
