Many polls and surveys testify to the apprehension older Americans have about their financial prospects in retirement. The lead article in this issue of The Journal of Retirement (JOR) speaks directly to these concerns. In “Strategies for Managing Retirement Risks,” David Laster, Nevenka Vrdoljak, and Anil Suri consider four risks to a financially secure retirement that are not simply due to a lack of self-control or poor planning: longevity risk, uncertain health care costs, sequence of returns risk, and inflation. Longevity risk, or an uncertain lifespan, raises the specter of running through all one’s money before death. It makes planning very difficult because knowing the average life expectancy for one’s age cohort does not make the date of one’s death predictable. The unpredictability of long-term care costs can pose a particular problem, given the expense of such care. Sequence of returns risk greatly complicates financial planning: Even if the average rate of return during retirement were predictable, variations in the sequence of returns can make a huge difference to the rate of sustainable spending from a nest egg of a given size. Finally, it is easy to overlook the erosive effects over time that even moderate rates of inflation have on the purchasing power of income.
The authors recommend four strategies to mitigate these risks. First, avoid automatically claiming Social Security at the earliest possible age unless health is problematic; delaying substantially increases the benefit and also mitigates inflation risk. Second, invest in a life annuity. These instruments address longevity risk since they do not stop paying during the annuitant’s life. Third, develop a plan for the systematic drawdown of income from a portfolio of assets (a systematic withdrawal program, or SWP). The share of assets it is safe to draw down increases with the age of the investor, with all other things equal. Finally, pre-retirees should plan seriously for the possible expense of long-term care. The authors rightly conclude that taken together these strategies can help mitigate some of the serious risks that retirees confront as they age.
The JOR has published quite a few articles on defined contribution (DC) plans, usually on best-practice investment strategies. However, DC plans raise very important regulatory or, more broadly, governance issues. In “Governance: The Sine Qua Non of Retirement Security,” Michael Drew and Adam Walk assess DC plan governance in Australia and find it wanting in certain respects. The basic similarity between the DC environments in Australia and the United States makes their article very relevant to the U.S. setting.
Their article starts with a vignette of two otherwise identical investors, one of whom liquidates his pension plan just before a market crash and the other just after. Despite the similarities of their financial positions, their retirement nest eggs differ drastically. This disparate experience, the authors maintain, should not happen.
The authors argue that a properly designed fiduciary focus is the key to retirement security. In particular, the focus should be on providing some given level of income in retirement. Regulation in Australia focuses on the accumulation phase, when it should be focused on what the authors describe as outputs from wealth, such as the level of sustainable retirement income.
Drew and Walk provide a checklist for good governance that includes providing the fiduciary entity with the time and resources necessary to oversee a plan’s operations adequately and a documented and comprehensive investment policy. They also set out a series of basic investment principles: Retirement income (not capital) should be the basic goal; investors are not all struck from the same mold, and their time horizons vary; and asset allocation should be the main concern of investment committees and should not be static. They also recommend various ways to guide investors and emphasize the importance of what they term implementation excellence and an effective risk management framework, even if DC plans have traditionally not emphasized this function. This well-written article raises a host of important issues.
In “Mandated Retirement Systems and Implied Benefits and Costs for Workers from Various Generations in Selected Countries,” Sylvester Schieber compares the key features and considers the prospects of the public pension systems of the United States, Canada, Australia, and the United Kingdom. His article touches on the macroeconomic aspects of national pension systems, such as their impact on saving rates and labor supply, but his main concern is with the returns in the form of pension benefits that workers can expect from their contributions. Specifically, he estimates money’s worth ratios (MWRs) for low earners, average earners, and high earners in each of the four systems based on the earnings histories of workers who have been employed for most of their careers. This analysis is preceded by a succinct summary of the basic features of each system. A basic version of the MWR can be expressed as the following:
where Pi stands for the pension benefit and Wi for the wage paid in year i; t for the payroll tax rate; pi for the probability of surviving until year i; r for the discount rate; R-1, for the last year of work; and T for the maximum retirement period. When the MWR equals one, the discounted value of the benefits a worker can expect to receive weighted by his or her survival probabilities equals the discounted value of the payroll tax contributions he or she has paid. Viewed as an investment, the worker’s return in this case exactly equals the discount rate.
MWRs are calculated for each country, an exercise that reveals some marked differences among them. Australia has the highest because its current system is basically a compulsory DC plan with a means-tested basic pension. The DC plan component will have an MWR of one, while the Age Pension, as it is called, being financed from general revenues, does not have a cost in this exercise. That said, poor investment performance could result in unacceptably low replacement ratios. In the United Kingdom, MWRs are quite high for the low paid. MWRs for the better paid U.K. workers exceed one in 2010 but are estimated to decline substantially over time. In Canada, MWRs are also comparatively high for the low paid—in all three countries, part of the pension benefit is financed from general revenues, which are not treated as costs by the MWR formula. The United States is the only country in which the entire benefit is earnings-related, which partly accounts for its relatively low value. MWRs for all three earnings classes are projected to decline over time, reflecting the projected decline in the number of workers per pensioner, or the increase in the dependency ratio.
The MWR projections for the United States, low as they are, are based on current law and assume no change in the parameters of the system, even in 2034 when the Trust Fund is estimated to be exhausted. Reforms can only worsen the U.S. numbers. The author emphasizes that Congress has yet to consider a single proposal that would put Social Security’s finances on a sustainable basis. The greater the extent to which current pensioners are safeguarded from the effects of reform, the greater the impact on the MWRs of younger workers.
The use of defaults to encourage participation and saving in 401(k) and other DC plans has been quite successful. However, a successful financial retirement depends not only on an adequate amount of saving but also on ensuring that the retirement nest egg will last throughout retirement. If default settings can be successful in encouraging participation and saving, the question naturally arises as to whether they could be applied to encourage annuitization of plan balances either before or at retirement. In “Defaulting Participants in Defined Contribution Plans into Annuities: Are the Potential Benefits Worth the Costs?” David Blanchett undertakes a thorough examination of this issue and basically concludes that the answer to his question is “yes.”
The author notes the reservations that plan providers have had regarding the provision of annuities, including fiduciary concerns and a lack of regulatory clarity, as well as the traditional reluctance of Americans to buy annuities. Nonetheless, the results of his stochastic modeling exercises (his model is described in some detail in a series of appendexes to the article) make it clear that a well-designed default setting that would result in the annuitization of some given percentage of a plan balance could be of great benefit to plan participants. At the same time, he stresses that the benefits of having an annuity vary greatly from one person to another.
In his model, he has included nine variables that encompass these personal differences, including the percentage of income that is guaranteed (like Social Security), the ages at which an annuity is purchased and at which income starts, subjective life expectancy, and what he terms the shortfall preference risk aversion factor. The role of guaranteed income is particularly important; a high level definitely reduces the need for an annuity. A policy of offering a default annuity may be most successful when plan members have relatively similar preferences and financial positions. In this case, it should be easier to gauge what the optimal rate of annuitization should be. A default annuity setting is likely to be more successful when plan participants lack a defined benefit plan, have a higher life expectancy (which tends to be related to income), are relatively risk averse, include a higher proportion of females (women benefit from gender-neutral pricing), are not particularly well prepared for retirement, and are generally similar to one another.
The financial security in retirement of the country’s veterans should be a chief concern of policy. How well prepared veterans are for retirement is not very well understood. In “Declining Wealth and Work among Male Veterans in the Health and Retirement Study,” Alan Gustman, Thomas Steinmeier, and Nahid Tabatabai explore this issue by comparing the financial and demographic status of veterans and nonveterans and its evolution over time. Specifically, they compare four cohorts from the Health and Retirement Study (HRS), a longitudinal study of older Americans. All four cohorts entered the study at ages 51–56: The original HRS cohort entered the study in 1992; the War Babies in 1998; the Early Boomers in 2004, and the Mid-Boomers in 2010. The data the HRS collects permit a comparison of basic demographic and financial indicators, and interesting trends emerge. Education indicators of nonveterans and veterans converge: Veterans from the original HRS cohort have on average one more year of schooling than nonveterans, but this gap is virtually eliminated by the time the Mid-Boomers enter the study, as is a gap between veterans and nonveterans concerning the population share who have at least some college. The incidence of health problems of nonveterans are little changed, but those of veterans take an alarming leap upward.
Similar trends are evident with wealth indicators. Total wealth of veterans from the original HRS cohorts averages $883,000 in 2010 dollars, compared to $822,000 for nonveterans; the wealth of veterans from the Mid-Boomers cohort declines to $648,000, while the wealth of nonveterans reaches $865,000. The present value of pensions paid to veterans declines drastically. In general, the components of personal wealth all decline in value but more so for veterans than nonveterans.
A regression analysis finds that the gap in wealth in favor of veterans from the original cohort can be explained by demographics; the gap in favor of nonveterans from the youngest cohort cannot. The authors conclude that veterans from the two oldest cohorts are (or were) better prepared for retirement than veterans from the other two. They speculate that the postdraft entrants into the armed services had weaker labor-market prospects than those of the general population, although the reasons for this discrepancy are hard to pin down.
In “An Approach to Financial Planning of Retirement Pensions with Scenario-Dependent Correlation Matrixes and Convex Risk Measures,” William Ziemba sets out a very sophisticated model that can be used by financial institutions to guide the management of retirement pensions. The model is based on the technique of stochastic linear programming and was developed by him and a colleague for a large Austrian defined contribution pension. This is quite a technical article, and the interested reader will want to read it carefully to understand how the model works. It has a number of innovative features, perhaps most notably its adoption of changing correlations across asset classes over time, which are made a function of overall market volatility.
The model is applied to a world of four asset classes: U.S. stocks, U.S. bonds, Euro stocks, and Euro bonds. Asset returns and correlations are derived from monthly data from January 1970 to September 2000 (from 1986 for bonds). This period is divided into three regimes: A period of normal volatility observed 70% of the time; a period of high volatility observed 20% of the time; and a period of extreme volatility observed 10% of the time. Cross-correlations vary substantially across these periods, notably that between Euro stocks and bonds.
The author explains that the model was simulated using both “mixing” correlations—where a switch of regimes can take place—and unchanged correlations initially over five periods totaling 10 years. Simulations assuming mixed correlations generally outperformed simulations assuming constant correlations. In particular, simulations through the model’s last period that assumed mixing correlations resulted in higher levels of terminal wealth and lower probabilities of a shortfall from the model’s assumed wealth target. The share of stocks tended to increase under a mixing correlation regime. In addition, when the actual wealth generated by the model fell substantially short of the pension plan’s wealth target, the need to avoid the shortfall pushed up the share of equities in the portfolio. A test of the model using monthly data from January 1992 to January 2002 with monthly rebalancing determined in part by the current volatility regime substantially outperformed a strategy that assumed no change in correlations.
Ethical issues, and notably the role of fiduciary standards, are a hot topic in the financial industry these days, and they and related concerns arise in the business of retirement planning as much as they do elsewhere. In “Ethical Issues in Retirement Income Planning: An Advisor’s Perspective,” Jamie Hopkins, Julie Ragatz, and Chuck Galli analyze the results of a survey they conducted among a group of retirement planning professionals. The article describes how the survey was organized and conducted while addressing some possible shortcomings with its design, including a low overall response rate.
In response to a series of about 50 questions dealing quite comprehensively with possible worries or concerns retirement advisors might have, the survey subjects reported that among their major concerns were the possibility of elder abuse; retirement professionals’ lack of understanding of Social Security, Medicare, and tax planning; clients’ ignorance of the complexity of retirement planning and the products they might be asked to buy; and sales representatives’ misrepresentation of the products they were recommending to their clients.
In general, the survey participants were not unduly concerned about the ethical climate of the industry and were skeptical of formal programs of ethics training. In their view, ethics cannot be taught; the corporate culture was more important. That said, they thought that a better-educated cadre of advisors would improve the ethical climate. Participants were concerned that their younger colleagues faced pressure to make sales and sell new products, even if these products did not really meet their clients’ needs. Notwithstanding their concern over elder abuse, participants were not aware of many acts of outright fraud. The survey also revealed concerns over the lack of clarity as to the requirements of a universal fiduciary standard.
In their concluding section, the authors state that “the dominant narrative of this study was concern over the lack of education.” A better-educated workforce would provide better and more conscientious advice. Enhancing consumer (client) literacy would help achieve the same end.
The dominant role played by the retirement component of Social Security (the “old-age” in Old-Age, Survivors, and Disability Insurance, or OASDI) can lead us to overlook the importance of the disability component, which currently pays pensions to about nine million Americans. In “Enhancing U.S. Retirement Security through Coordinated Reform of Social Security Disability and Retirement Insurance Programs,” Jason Fichtner and Jason Seligman provide us with a comprehensive overview of the financial position and prospects of Social Security Disability Insurance (SSDI) and make the case for reforms that would enhance the welfare of many disabled persons. The main theme of their article is the need to coordinate reforms of SSDI with the rest of the system. As a simple example, measures that reduce the old-age pension will increase applications for the disability pension and so may not have the anticipated impact on the finances of the system as a whole. Although merging the trust funds for Old-Age and Disability Insurance could push back the day when SSDI funds are exhausted, it is not a fundamental solution to SSDI’s financial problems.
SSDI currently restricts benefits to those who are permanently and totally disabled. This stance is reflected in its policy toward those of the disabled who can work: Once monthly earnings exceed about $1,090, the benefit risks being completely rescinded. The authors argue that policy should be more flexible and should offer benefits to those affected by partial and temporary disability. They also note that the SSDI program is not well coordinated with the Americans with Disabilities Act (ADA), perhaps because SSDI is primarily financial, while the ADA is a set of regulations.
Fichtner and Seligman also propose that the private disability insurance industry should play a greater role. For example, employers could be required to buy group private insurance and to cover the first two years of an employee’s disability. This would increase the incentive employers would have to keep partially disabled employees on the job and take action to make the workplace safer. However, additional rules would be necessary to avoid encouraging employers to game the system—for example, by not hiring workers from groups whose members are prone to disability. The reader is invited to explore these and other reform proposals in this very interesting article.
TOPICS: Retirement, tail risks, social security, legal/regulatory/public policy
George A. (Sandy) Mackenzie
Editor
Footnotes
Publisher’s Note
Institutional Investor, the publisher of The Journal of Retirement, wants to extend a special thank you to Bank of America Merrill Lynch for its support of The Journal of Retirement. Please note that no sponsor has influence on the editorial content found in The Journal of Retirement. Representatives from any firm are encouraged to submit an article to our independent Editor, Sandy Mackenzie, for review and prospective acceptance into the publication. All editorial submissions, acceptance, and revisions are the sole decision of Mr. Mackenzie. The editorial submission guidelines are found on the last page of the publication and online at www.iijor.com. Thank you, and I hope you enjoy this and future issues of The Journal of Retirement.
Dave Blide
Publisher, Institutional Investor Journals, dblide{at}iijournals.com
- © 2016 Pageant Media Ltd