A Retirement Adequacy Analysis of Default Options ... - Nxtbook Media
䉴 Default Options and Lifecycle Funds
A Retirement Adequacy Analysis of Default Options and Lifecycle Funds by Jason Ellement and Lori Lucas
Retirement income replacement analysis can assist defined contribution plan sponsors in selecting default investment. Using quantitative analysis, this article demonstrates that the plan sponsor’s default investment selection has a direct and important impact on employees’ retirement income adequacy. The authors also demonstrate the importance of other automatic enrollment program design elements, such as the implication of different lifecycle paths for postretirement asset allocation and the importance of robust default contribution and escalation rates.
he 2006 Pension Protection Act (PPA) ushered in a new era of benign paternalism toward defined contribution (DC) plans. PPA encourages the use of such features as automatic enrollment, contribution escalation and diversified default investment options. The goal of automating DC plans is to facilitate greater retirement income adequacy for employees at a time when U.S. workers must increasingly rely on their DC plan for a secure financial future. For plan sponsors, implementing the “right” default options is both challenging and absolutely critical. Plan sponsors have learned that DC participants tend to remain in defaults for prolonged periods of time. The composition of plan defaults, therefore, has a direct and important impact on employees’ retirement income adequacy. In this article, the suitability of various default investment options is examined. The authors consider three possible default options: A principal preservation fund, a balanced fund and a lifecycle fund. In issuing its initial guidelines for qualified default
T 34
BENEFITS QUARTERLY, Third Quarter 2007
investment alternatives (QDIAs), which would receive safe harbor protection under ERISA Section 404(c) according to PPA, the Department of Labor (DOL) specified that balanced funds, target date (lifecycle funds) and managed accounts qualify as QDIAs, while stable value funds do not. The authors define suitability as the impact of default investment choices on potential retirement income adequacy. Retirement income adequacy is measured by simulating income replacement ratios under a multitude of economic scenarios.1 In terms of providing retirement adequacy, this article will demonstrate that a well-designed lifecycle fund is a superior default option. Specifically, the variety in lifecycle funds and how intelligent asset allocation design can enhance retirement adequacy is examined. Throughout this article, empirical results are based on the following assumptions, unless otherwise stated: A participant joins a 401(k) plan at the age of 25 and begins contributing 6% of pay annually, with an employer match of 50% (total contribution of 9%). The participant’s starting salary is initially $25,000, and future annual salary increases are simulated (expected salary increase is 3.5% per year). These assumptions
䉴 FIGURE 1
LIFECYCLE PATH
reflect general U.S. population demographics and the long-term nature of a 401(k) plan. Many U.S. workers do not start saving for retirement at the age of 25. Nonetheless, default investment options and lifecycle fund construction should be evaluated in the context of long-term saving. Not saving enough—either the level of saving or length of saving period—is a funding problem not within the scope of this article, although the authors do illustrate the impact that a higher or lower savings level can have on retirement adequacy.
THE ADEQUACY OF DEFAULT INVESTMENT OPTIONS First, consider the retirement income adequacy of three default investment options: A principal preservation fund that is 100% stable value, a balanced fund comprised of 60% large-cap equity (S&P 500) and 40% U.S. fixed income (Lehman Aggregate), and a lifecycle fund (target date fund). A lifecycle fund is an asset allocation path that changes over time. With this in mind, the authors refer to a lifecycle fund as a lifecycle path, or glide path, for the remainder of this article. A lifecycle path is shown in Figure 1. The underlying premise of all lifecycle paths is to
䉴 THE AUTHORS Jason Ellement is a consultant and shareholder with Capital Markets Research Group. He is responsible for assisting plan sponsor clients with strategic planning, conducting asset and liability studies, developing optimal investment manager structures and providing custom research on a variety of investment topics. Ellement is a fellow of the Society of Actuaries, a chartered financial analyst and a member of the Academy of Actuaries and Security Analysts Society of San Francisco. He has a bachelor of science degree in actuarial science and statistics from the University of Manitoba, Canada. Lori Lucas, CFA, is senior vice president and the DC practice leader of Callan Associates in Chicago. She is responsible for setting the direction of Callan’s DC business, providing DC support both internally to Callan’s consultants and externally to Callan’s clients, and developing research and insight into DC trends. Lucas received a bachelor of arts degree from Indiana University and earned a master’s degree from the University of Illinois.
BENEFITS QUARTERLY, Third Quarter 2007
35
䉴 FIGURE 2
PROJECTED ASSETS (TODAY’S DOLLARS)
Percentile
Principal Preservation Fund
5th (Best Case) 25th 50th (Expected Case) 75th 95th (Worse Case)
$176,000 $162,000 $152,000 $145,000 $133,000
reduce portfolio risk (typically a reduction in equity exposure) as the individual’s time horizon shortens. Young investors, with long time horizons, have ample time to recover from large market losses and are making large contributions (relative to the size of their account balance). In addition to the many “earning years” in front of the young investor, the capacity to take investment risk is very high. In the years eligible for retirement (ages 50 to 65) the ability to absorb risk gradually declines as preservation of retirement wealth becomes more important. Eligible retirees have fewer “earning years” in front of them. Upon retirement, many glide paths continue to roll down equity exposure as the individual’s time horizon (life span) gets shorter. The above lifecycle path is generic and comprised of only two asset classes. Diversification and customization of the lifecycle path are discussed later in this article. In Figure 2, the authors compare the range of projected asset values (expected case, as well as various best- and worst-case scenarios) for each of the three possible default investment options. Assets
36
BENEFITS QUARTERLY, Third Quarter 2007
Balanced Fund
Lifecycle Path
$628,000 $396,000 $285,000 $213,000 $136,000
$755,000 $457,000 $317,000 $226,000 $138,000
are accumulated from ages 25 to 65 and are adjusted for inflation, so asset values are displayed in today’s dollars. Over a long time horizon, the opportunity cost of forgoing capital appreciation is enormous. Over 40 years, the expected values of both the balanced fund and the lifecycle path dramatically surpass that of the principal preservation fund. Indeed, the expected value of the lifecycle path is more than twice that of the principal preservation fund. This analysis demonstrates that a principal preservation fund does not serve DC participants well as a default investment option for a long-term funding retirement vehicle. The analysis in Figure 2 also shows that the lifecycle path offers more long-term growth potential than the balanced fund ($317,000 versus $285,000 in the expected case). Even the worst-case scenario (95th percentile) for the lifecycle path is slightly higher than that of the balanced fund. The lifecycle path achieves more asset accumulation over the 40-year horizon through more intelligent asset allocation design. A
䉴 FIGURE 3
INCOME REPLACEMENT RATIO AT AGE 65
Percentile
Principal Preservation Fund
5th (Best Case) 33% 25th 31% 50th (Expected Case) 29% 75th 27% 95th (Worse Case) 25% Probability 0% (replacement ratio ⬎ 60%) 60%/40% balanced fund is too conservative for young investors and too aggressive for those close to retirement. The lifecycle path, on the other hand, focuses on asset growth for many years prior to the age of 50 and then begins to emphasize principal preservation as retirement approaches. Projected asset values do not necessarily convey retirement adequacy. Is $317,000 sufficient to maintain a participant’s standard of living in retirement? To answer this question, the authors convert the projected asset value of each default investment option to an annual life income stream in order to calculate an income replacement ratio.2 An income replacement ratio is simply annual income in retirement divided by final salary prior to retirement. If a participant’s final salary before retirement is $100,000, and the annuitization of the participant’s account balance provides $60,000 in the first year, the income replacement ratio is 60%.3 Figure 3 depicts the income replacement ratios for the three default investment options. The last row in
Balanced Fund
Lifecycle Path
119% 75% 54% 40% 26% 41%
143% 86% 60% 43% 26% 50%
the table is the probability of the projected replacement ratio exceeding 60% at the age of 65. The income replacement ratios reflect income from the 401(k) plan only and do not include Social Security benefits. Social Security may add another 30% income replacement for this hypothetical participant.4 Research has suggested that individuals should target no less than 75-90% income replacement from all sources to retire adequately—and even more to accommodate postretirement medical costs.5 When asked, most workers believe that they need to replace no more than 85% of income, with half believing that they can live adequately in retirement on 70% or less of preretirement income.6 Although the appropriate level of income replacement depends on many factors—income level, gender, health, net worth—the authors assume a 90% income replacement rate is necessary for this study. Given Social Security replacement of 30%, the 401(k) plan therefore needs to provide 60% income replacement. The probability of achieving a 60% replacement rate
BENEFITS QUARTERLY, Third Quarter 2007
37
䉴 FIGURE 4
SAMPLE OF EIGHT MUTUAL FUND LIFECYCLE PATHS
through the 401(k) plan is therefore a reasonable metric to measure the potential retirement adequacy of an investment strategy. The principal preservation fund is expected to replace only 29% of preretirement income and has no chance of achieving a 60% replacement rate. In contrast, the balanced fund and the lifecycle path are expected to replace 54% and 60%, respectively, with a much higher probability of meeting the 60% target. The balanced fund and lifecycle path have a 41% and 50% probability of meeting the 60% replacement target. Of course, a financially secure retirement cannot be supported by a 50-50 chance of meeting an individual’s retirement objective. Now, we turn our attention to design and customization issues related to lifecycle paths in an effort to improve the chances of meeting a 60% replacement rate.
LIFECYCLE PATHS ARE NOT CREATED EQUAL First, consider the equity exposure of a lifecycle path. Figure 4 shows the equity exposure of eight well-known lifecycle mutual funds.
38
BENEFITS QUARTERLY, Third Quarter 2007
In Figure 4, the terminal equity allocation (age 95) varies from 17-44%. On the opposite end of the spectrum, the initial equity allocation varies from 81% to 93%. Interestingly, the largest disparity of equity allocation is at the age of 65. Here, the equity allocation varies from 28% to 58%, a range of 30%. In other words, the appropriate asset allocation going into the age of 65 (retirement) is the greatest area of debate regarding lifecycle paths. The wide dispersion in equity allocation across a small sample of off-the-shelf products suggests that plan sponsors are afforded considerable choice in the market and should exercise due diligence in the selection (or construction) of lifecycle paths. In order to demonstrate the impact of different levels of equity exposure on retirement adequacy, the authors construct and analyze three simplified lifecycle paths comprised of two asset classes: large cap U.S. equity and U.S. fixed income.The lifecycle paths represent a conservative, moderate and aggressive glide path.The moderate lifecycle path consists of the same asset allocation path analyzed in Figure 1. The equity exposure of all three lifecycle paths is shown in Figure 5. Figure 6 shows the income replacement ratio at the age of 65 of each of the glide paths.
䉴 FIGURE 5
EQUITY ALLOCATION OF LIFECYCLE PATHS
The expected replacement ratio ranges from 5265%, while the probability of meeting the 60% replacement target ranges from 40-55%. As expected, equity variation among lifecycle paths results in wide dispersion of replacement ratios. Greater dispersion in replacement rates is observed on the upside (25th and 5th percentiles) rather than the downside. Over a long time horizon with the compounding of returns, a more aggressive lifecycle path has significantly more upside potential. On the downside (95th percentile), the worst-case replacement ratio for the aggressive path is no worse than the conservative and moderate paths. The aggressive path may be more volatile than the other paths, but the long time horizon leads to greater asset accumulation. The aggressive path can achieve a replacement ratio similar to that of the conservative path in the worst-case scenario simply because it has more assets to lose when disaster strikes. An aggressive lifecycle path is attractive from a long-term perspective but would have high short-term volatility during the eligible retirement period. Individuals nearing retirement become more sensitive to short-term volatility and may prefer a more conserva-
tive lifecycle path from ages 50 to 65. Figure 7 shows the one-year percentage decline in assets in a worstcase scenario.7 A potential one-year decline (5% chance) in assets of 9-12% from ages 50 to 65 may be too high for some investors. Plan sponsors may wish to take shortterm volatility into consideration as they evaluate the appropriateness of glide paths. Plan sponsors may find that younger participants are also sensitive to short-term volatility. Research on myopic risk aversion suggests that a long-term investment horizon associated with retirement does not necessarily mean that participants will be insensitive to near-term losses.8 Plan sponsors that are concerned about their participants’ ability to stay the course during market downturns may wish to weigh interim volatility more heavily than those that believe their DC participants will have a long-term focus. Plan sponsors may wish to consider employee turnover when selecting the aggressiveness of the glide path. According to a Hewitt Associates study, 45% of 401(k) participants take a cash distribution upon cessation of employment (this number is even higher for younger workers).9 If cashouts are likely,
BENEFITS QUARTERLY, Third Quarter 2007
39
䉴 FIGURE 6
INCOME REPLACEMENT RATIO AT AGE 65
Percentile
Conservative Path
5th (Best Case) 111% 25th 71% 50th (Expected Case) 52% 75th 39% 95th (Worse Case) 26% Probability 40% (replacement ratio ⬎ 60%)
short-term volatility may figure more prominently in the glide path selection equation. Fortunately, employees have many options for preserving the tax-deferred status of their retirement assets upon cessation of employment. Former employees may be able to leave assets in their old 401(k) plan or roll over to another tax-deferred retirement vehicle, such as the successor employer defined contribution plan or an individual retirement account (IRA). A lifecycle path should be developed with the primary goal of retirement income adequacy. Also note, in Figure 7 the lifecycle paths are projected to experience smaller and smaller possible one-year declines in assets over time, as their equity allocation declines. Because the balanced fund maintains a constant level of equity exposure, this leads to a perverse increase in risk as retirement approaches. The potential for a larger one-year decline in assets under the balanced fund is a result of assets trending
40
BENEFITS QUARTERLY, Third Quarter 2007
Moderate Path
Aggressive Path
143% 86% 60% 43% 26% 50%
171% 97% 65% 44% 26% 55%
upward under a constant level of equity exposure. The balanced fund may be too aggressive for participants over the age of 60.
OTHER ASSET CLASSES AND DIVERSIFICATION Another possible way to improve retirement adequacy is through diversification. So far, our lifecycle paths have only consisted of two asset classes. Figure 8 introduces six new asset classes to the moderate lifecycle path. In Figure 8, the moderate path’s equity exposure is diversified into small/mid cap U.S. equity and nonU.S. equity. Not only does this improve overall diversification, the potential for adding excess return, or return above the benchmark, in small/mid cap U.S. equity and non-U.S. equity is higher than for large cap U.S. equity. In the context of the overall portfolio, other ef-
䉴 FIGURE 7
ONE-YEAR PERCENT DECLINE IN ASSETS WORST-CASE SCENARIO (95TH PERCENTILE)
fective diversifiers include real estate investment trusts (REITs), treasury inflation-protected securities (TIPS) and high-yield bonds. As hybrid securities, REITs and high-yield bonds have risk/return characteristics that are between equity and U.S. fixed income and can enhance or maintain asset growth if funded from a combination of equities and U.S. fixed income. The allocation to REITs and the ratio of high-yield bonds to U.S. fixed income is reduced from age 50 onward to dampen portfolio volatility. TIPS are not a growth asset but rather serve as an effective hedge against both expected and unanticipated inflation. Equities are an effective longterm hedge against inflation, but the equity exposure in the glide path begins to decline dramatically from age 50 onward. A participant is most vulnerable to inflation from ages 50 to 70, when the participant has built a large account balance due to years of prefunding and investing. TIPS are also a lowrisk asset (low volatility and weak correlations to other asset classes) and can be used to reduce glide
path volatility. The value of TIPS to older participants results in an allocation that emerges at the age of 50 and gradually grows as a percent of the portfolio thereafter. Short-duration bonds are introduced after the age of 50 to further dampen portfolio volatility. The allocation to short-duration bonds increases as retirement approaches and the participant’s time horizon (life span) shortens. Figure 9 shows the range of income replacement ratios for the moderate-diversified path compared to that of the nondiversified moderate and aggressive paths. The moderate-diversified path is expected to replace 7% and 12% more income than the nondiversified aggressive and moderate paths, respectively. The probability of achieving the 60% replacement target increases 7% to 12% by moving to the moderatediversified path as well. Most impressive of all, the moderate-diversified path has significantly more upside potential (5th percentile) and a higher replacement ratio in the worst-case scenario. Clearly,
BENEFITS QUARTERLY, Third Quarter 2007
41
䉴 FIGURE 8
MODERATE-DIVERSIFIED LIFECYCLE PATH
diversification and thoughtful asset allocation can significantly improve a lifecycle path.
ALTERNATIVE INVESTMENTS As lifecycle paths begin to push the envelope as far as employing alternative investments within the DC plan structure, they face challenges such as the need for daily valuation and daily liquidity. For example, private equity programs involve a 15-20 year time horizon with capital calls and net distributions at periods set by the manager, not the investor. Likewise, hedge funds typically have lockup periods and limited opportunities for contributions and redemptions. Finally, private real estate is subject to infrequent appraisal smoothing. Currently, a great deal of energy and time is being expended by product innovators and other interested parties to incorporate illiquid asset classes (or a close representation) into a DC plan structure. In the near future, it may be possible to include traditionally illiquid asset classes in the lifecycle path. Plan sponsors, especially those that are building lifecycle paths out of their underlying core investment options, should stay alert to any developments.
42
BENEFITS QUARTERLY, Third Quarter 2007
POSTRETIREMENT ASSET ALLOCATION The event of retirement is an important inflection point in the glide path, as the investor shifts from an accumulation phase to a spending phase. At this stage of the lifecycle, the risk of outliving assets (longevity risk) and the erosion of purchasing power due to inflation are primary threats to future financial security. A significant allocation to assets with capital growth potential may thwart these risks. At the same time, the ability to assume market risk is more limited for a retiree. Unlike the situation of the young investor with an ample recovery period and assets flowing into the 401(k) plan, a retiree is spending down the assets. Meanwhile, a contracting life expectancy leads one to question the sufficiency of the recovery period. A severe market downturn that coincides with spending down assets can deplete a participant’s balance to such a low level that an older retiree may not have sufficient time to recoup investment losses. Figure 10 shows the probability of depleting the participant’s balance to zero (starting at age 65 retirement) by ages 70 to 105 years for four lifecycle paths
䉴 FIGURE 9
INCOME REPLACEMENT RATIO AT AGE 65
Percentile
Moderate Path
5th (Best Case) 143% 25th 86% 50th (Expected Case) 60% 75th 43% 95th (Worse Case) 26% Probability 50% (replacement ratio ⬎ 60%) examined earlier. In this analysis, the participant is assumed to spend 60% of final salary each year directly from the 401(k) plan. In addition, spending is assumed to increase each year with simulated inflation to keep pace with a rising cost of living. As expected, the probability of depleting assets to zero rises as a participant gets older. The life expectancy of someone aged 65 is 18 years (the age of 83).10 In other words, 50% of participants are expected to live beyond the age of 83. The moderatediversified path stands out, since it has the lowest probability of asset depletion and with less short-term volatility than the aggressive path. Interestingly, the conservative path is the most risky, as the probability of asset depletion is the highest. This is in stark contrast to Figure 7 where the conservative path was the low-risk option during ages 50 to 65. In Figure 7, risk was defined as a oneyear decline in assets (a short-term measure of risk). The cost of lower short-term volatility is lower asset growth prior to the age of 65, resulting in a higher
Aggressive Path
ModerateDiversified Path
171% 97% 65% 44% 26% 55%
197% 105% 72% 49% 29% 62%
probability of depleting assets prematurely. Although low volatility may be desirable as retirement approaches, investors have to keep in mind that they may live another 20-30 years beyond the age of 65.
SAVINGS RATES The analysis thus far has assumed contribution levels of 9% of pay (6% employee and 3% employer contributions) until the age of 65. In this section, the authors contrast the retirement income adequacy achieved with a 5%, 9% and 13% savings policy. The moderate-diversified (“Mod-Div”) lifecycle path is the underlying default investment option in each case. Once again, the range of replacement rates is shown in Figure 11. The level of savings clearly has a dramatic impact on retirement adequacy. The probability of achieving a 60% replacement rate increases from 23% to 62% when contributions are increased from 5% of pay to 9% of pay. A further increase of 4% (from 9% of pay
BENEFITS QUARTERLY, Third Quarter 2007
43
䉴 FIGURE 10
PROBABILITY OF ASSET DEPLETION
to 13% of pay) increases the probability of hitting the 60% target from 62% to 83%. Increasing contributions also shifts the entire range of probable replacement ratios upward leading to less downside risk (higher replacement in the 95th percentile) and significantly more upside potential. Figure 11 is an important reminder that a robust default contribution rate and the use of automatic contribution escalation are critical factors to consider when developing automatic enrollment programs.
CONCLUSION As the U.S. retirement system moves from a defined benefit to a defined contribution focus, the authors find a simultaneous evolution of defined contribution plans incorporating many of the features traditionally associated with defined benefit plans. Automatic pension plan participation is being mirrored by automatic enrollment in the DC environment. Payouts that increase with pay in the DB world are mirrored with automatic contribution escalation in the DC environment.
44
BENEFITS QUARTERLY, Third Quarter 2007
By using retirement income replacement analysis as a means of evaluating default investment options, plan sponsors can take a page from the asset/liability type analysis that has long been a cornerstone of pension plan management. Just as asset/liability analysis seeks to match an appropriate asset allocation with pension funding needs, retirement income replacement analysis seeks to provide valuable insight into how well a series of lifecycle funds will meet defined contribution plan participants’ retirement funding needs. Of course, there are additional factors that need to be addressed in selecting an appropriate default option: the risk tolerance and demographics of the employee population, the degree of employer paternalism and the existence of other retirement vehicles to support and enhance retirement adequacy. One thing is clear, however: The days of assuming that the impact on retirement income adequacy is negligible when it comes to the default option are long gone. Going forward, the selection of the default investment option will be one of the most critical decisions that plan sponsors make when designing their defined contribution plan. 䉳
䉴 FIGURE 11
INCOME REPLACEMENT RATIO AT AGE 65
Percentile
ModerateDiversified (5% of Pay)
5th (Best Case) 110% 25th 59% 50th (Expected Case) 40% 75th 27% 95th (Worse Case) 16% Probability 23% (replacement ratio ⬎ 60%)
Endnotes 1. Callan’s proprietary Monte Carlo simulation model is employed to simulate investment returns and inflation across 1,000 economic scenarios. 2. To determine income replacement ratios, the account balance must first be converted to a life income stream. Annuity assumptions are a 5% interest rate, RP-2000 mortality with 50/50 blend (male/female) and 2.75% contractual indexing. (Inflation is expected to be 2.75% per year.) 3. Assumed interest rate for annuity conversion is 5% across all simulated scenarios. In a higher (lower) interest rate environment, annuities are cheaper (more expensive) and lead to higher (lower) replacement rates. 4. Hewitt Associates’ 2005 Research Report: Total Retirement Income at Large Companies: The Real Deal.
ModerateDiversified (9% of Pay)
ModerateDiversified (13% of Pay)
197% 105% 72% 49% 29% 62%
285% 152% 104% 71% 41% 83%
5. Measuring Retirement Income Adequacy: Calculating Realistic Income Replacement Rates. Jack VanDerhei. EBRI Issue Brief, No. 297, September 2006. 6. The Employee Benefit Research Institute’s 2006 Retirement Confidence Survey. 7. A linear line is fitted to each year’s simulated worst-case decline in assets by least squares regression. 8. “The Effect of Myopia and Loss Aversion on Risk Taking: An Experimental Test.” Richard Thaler, Amos Tversky, Daniel Kahneman and Alan Schwartz. Quarterly Journal of Economics. Volume 112, pages: 647-61. 9. “Are Workers Preserving Their 401(k) Retirement Wealth?” Hewitt Associates. July 2005. 10. Based on RP-2000 Mortality table published by the Society of Actuaries. Mortality rates are blended 50/50 male/female.
International Society of Certified Employee Benefit Specialists Reproduced from the Third Quarter 2007 issue of BENEFITS QUARTERLY, published by the International Society of Certified Employee Benefit Specialists. With the exception of official Society announcements, the opinions given in articles are those of the authors. The International Society of Certified Employee Benefit Specialists disclaims responsibility for views expressed and statements made in articles published. No further transmission or electronic distribution of this material is permitted without permission. Subscription information can be found at www.iscebs.org. ©2007 International Society of Certified Employee Benefit Specialists BENEFITS QUARTERLY, Third Quarter 2007
45
A Retirement Adequacy Analysis of Default Options and Lifecycle Funds by Jason Ellement and Lori Lucas
Retirement income replacement analysis can assist defined contribution plan sponsors in selecting default investment. Using quantitative analysis, this article demonstrates that the plan sponsor’s default investment selection has a direct and important impact on employees’ retirement income adequacy. The authors also demonstrate the importance of other automatic enrollment program design elements, such as the implication of different lifecycle paths for postretirement asset allocation and the importance of robust default contribution and escalation rates.
he 2006 Pension Protection Act (PPA) ushered in a new era of benign paternalism toward defined contribution (DC) plans. PPA encourages the use of such features as automatic enrollment, contribution escalation and diversified default investment options. The goal of automating DC plans is to facilitate greater retirement income adequacy for employees at a time when U.S. workers must increasingly rely on their DC plan for a secure financial future. For plan sponsors, implementing the “right” default options is both challenging and absolutely critical. Plan sponsors have learned that DC participants tend to remain in defaults for prolonged periods of time. The composition of plan defaults, therefore, has a direct and important impact on employees’ retirement income adequacy. In this article, the suitability of various default investment options is examined. The authors consider three possible default options: A principal preservation fund, a balanced fund and a lifecycle fund. In issuing its initial guidelines for qualified default
T 34
BENEFITS QUARTERLY, Third Quarter 2007
investment alternatives (QDIAs), which would receive safe harbor protection under ERISA Section 404(c) according to PPA, the Department of Labor (DOL) specified that balanced funds, target date (lifecycle funds) and managed accounts qualify as QDIAs, while stable value funds do not. The authors define suitability as the impact of default investment choices on potential retirement income adequacy. Retirement income adequacy is measured by simulating income replacement ratios under a multitude of economic scenarios.1 In terms of providing retirement adequacy, this article will demonstrate that a well-designed lifecycle fund is a superior default option. Specifically, the variety in lifecycle funds and how intelligent asset allocation design can enhance retirement adequacy is examined. Throughout this article, empirical results are based on the following assumptions, unless otherwise stated: A participant joins a 401(k) plan at the age of 25 and begins contributing 6% of pay annually, with an employer match of 50% (total contribution of 9%). The participant’s starting salary is initially $25,000, and future annual salary increases are simulated (expected salary increase is 3.5% per year). These assumptions
䉴 FIGURE 1
LIFECYCLE PATH
reflect general U.S. population demographics and the long-term nature of a 401(k) plan. Many U.S. workers do not start saving for retirement at the age of 25. Nonetheless, default investment options and lifecycle fund construction should be evaluated in the context of long-term saving. Not saving enough—either the level of saving or length of saving period—is a funding problem not within the scope of this article, although the authors do illustrate the impact that a higher or lower savings level can have on retirement adequacy.
THE ADEQUACY OF DEFAULT INVESTMENT OPTIONS First, consider the retirement income adequacy of three default investment options: A principal preservation fund that is 100% stable value, a balanced fund comprised of 60% large-cap equity (S&P 500) and 40% U.S. fixed income (Lehman Aggregate), and a lifecycle fund (target date fund). A lifecycle fund is an asset allocation path that changes over time. With this in mind, the authors refer to a lifecycle fund as a lifecycle path, or glide path, for the remainder of this article. A lifecycle path is shown in Figure 1. The underlying premise of all lifecycle paths is to
䉴 THE AUTHORS Jason Ellement is a consultant and shareholder with Capital Markets Research Group. He is responsible for assisting plan sponsor clients with strategic planning, conducting asset and liability studies, developing optimal investment manager structures and providing custom research on a variety of investment topics. Ellement is a fellow of the Society of Actuaries, a chartered financial analyst and a member of the Academy of Actuaries and Security Analysts Society of San Francisco. He has a bachelor of science degree in actuarial science and statistics from the University of Manitoba, Canada. Lori Lucas, CFA, is senior vice president and the DC practice leader of Callan Associates in Chicago. She is responsible for setting the direction of Callan’s DC business, providing DC support both internally to Callan’s consultants and externally to Callan’s clients, and developing research and insight into DC trends. Lucas received a bachelor of arts degree from Indiana University and earned a master’s degree from the University of Illinois.
BENEFITS QUARTERLY, Third Quarter 2007
35
䉴 FIGURE 2
PROJECTED ASSETS (TODAY’S DOLLARS)
Percentile
Principal Preservation Fund
5th (Best Case) 25th 50th (Expected Case) 75th 95th (Worse Case)
$176,000 $162,000 $152,000 $145,000 $133,000
reduce portfolio risk (typically a reduction in equity exposure) as the individual’s time horizon shortens. Young investors, with long time horizons, have ample time to recover from large market losses and are making large contributions (relative to the size of their account balance). In addition to the many “earning years” in front of the young investor, the capacity to take investment risk is very high. In the years eligible for retirement (ages 50 to 65) the ability to absorb risk gradually declines as preservation of retirement wealth becomes more important. Eligible retirees have fewer “earning years” in front of them. Upon retirement, many glide paths continue to roll down equity exposure as the individual’s time horizon (life span) gets shorter. The above lifecycle path is generic and comprised of only two asset classes. Diversification and customization of the lifecycle path are discussed later in this article. In Figure 2, the authors compare the range of projected asset values (expected case, as well as various best- and worst-case scenarios) for each of the three possible default investment options. Assets
36
BENEFITS QUARTERLY, Third Quarter 2007
Balanced Fund
Lifecycle Path
$628,000 $396,000 $285,000 $213,000 $136,000
$755,000 $457,000 $317,000 $226,000 $138,000
are accumulated from ages 25 to 65 and are adjusted for inflation, so asset values are displayed in today’s dollars. Over a long time horizon, the opportunity cost of forgoing capital appreciation is enormous. Over 40 years, the expected values of both the balanced fund and the lifecycle path dramatically surpass that of the principal preservation fund. Indeed, the expected value of the lifecycle path is more than twice that of the principal preservation fund. This analysis demonstrates that a principal preservation fund does not serve DC participants well as a default investment option for a long-term funding retirement vehicle. The analysis in Figure 2 also shows that the lifecycle path offers more long-term growth potential than the balanced fund ($317,000 versus $285,000 in the expected case). Even the worst-case scenario (95th percentile) for the lifecycle path is slightly higher than that of the balanced fund. The lifecycle path achieves more asset accumulation over the 40-year horizon through more intelligent asset allocation design. A
䉴 FIGURE 3
INCOME REPLACEMENT RATIO AT AGE 65
Percentile
Principal Preservation Fund
5th (Best Case) 33% 25th 31% 50th (Expected Case) 29% 75th 27% 95th (Worse Case) 25% Probability 0% (replacement ratio ⬎ 60%) 60%/40% balanced fund is too conservative for young investors and too aggressive for those close to retirement. The lifecycle path, on the other hand, focuses on asset growth for many years prior to the age of 50 and then begins to emphasize principal preservation as retirement approaches. Projected asset values do not necessarily convey retirement adequacy. Is $317,000 sufficient to maintain a participant’s standard of living in retirement? To answer this question, the authors convert the projected asset value of each default investment option to an annual life income stream in order to calculate an income replacement ratio.2 An income replacement ratio is simply annual income in retirement divided by final salary prior to retirement. If a participant’s final salary before retirement is $100,000, and the annuitization of the participant’s account balance provides $60,000 in the first year, the income replacement ratio is 60%.3 Figure 3 depicts the income replacement ratios for the three default investment options. The last row in
Balanced Fund
Lifecycle Path
119% 75% 54% 40% 26% 41%
143% 86% 60% 43% 26% 50%
the table is the probability of the projected replacement ratio exceeding 60% at the age of 65. The income replacement ratios reflect income from the 401(k) plan only and do not include Social Security benefits. Social Security may add another 30% income replacement for this hypothetical participant.4 Research has suggested that individuals should target no less than 75-90% income replacement from all sources to retire adequately—and even more to accommodate postretirement medical costs.5 When asked, most workers believe that they need to replace no more than 85% of income, with half believing that they can live adequately in retirement on 70% or less of preretirement income.6 Although the appropriate level of income replacement depends on many factors—income level, gender, health, net worth—the authors assume a 90% income replacement rate is necessary for this study. Given Social Security replacement of 30%, the 401(k) plan therefore needs to provide 60% income replacement. The probability of achieving a 60% replacement rate
BENEFITS QUARTERLY, Third Quarter 2007
37
䉴 FIGURE 4
SAMPLE OF EIGHT MUTUAL FUND LIFECYCLE PATHS
through the 401(k) plan is therefore a reasonable metric to measure the potential retirement adequacy of an investment strategy. The principal preservation fund is expected to replace only 29% of preretirement income and has no chance of achieving a 60% replacement rate. In contrast, the balanced fund and the lifecycle path are expected to replace 54% and 60%, respectively, with a much higher probability of meeting the 60% target. The balanced fund and lifecycle path have a 41% and 50% probability of meeting the 60% replacement target. Of course, a financially secure retirement cannot be supported by a 50-50 chance of meeting an individual’s retirement objective. Now, we turn our attention to design and customization issues related to lifecycle paths in an effort to improve the chances of meeting a 60% replacement rate.
LIFECYCLE PATHS ARE NOT CREATED EQUAL First, consider the equity exposure of a lifecycle path. Figure 4 shows the equity exposure of eight well-known lifecycle mutual funds.
38
BENEFITS QUARTERLY, Third Quarter 2007
In Figure 4, the terminal equity allocation (age 95) varies from 17-44%. On the opposite end of the spectrum, the initial equity allocation varies from 81% to 93%. Interestingly, the largest disparity of equity allocation is at the age of 65. Here, the equity allocation varies from 28% to 58%, a range of 30%. In other words, the appropriate asset allocation going into the age of 65 (retirement) is the greatest area of debate regarding lifecycle paths. The wide dispersion in equity allocation across a small sample of off-the-shelf products suggests that plan sponsors are afforded considerable choice in the market and should exercise due diligence in the selection (or construction) of lifecycle paths. In order to demonstrate the impact of different levels of equity exposure on retirement adequacy, the authors construct and analyze three simplified lifecycle paths comprised of two asset classes: large cap U.S. equity and U.S. fixed income.The lifecycle paths represent a conservative, moderate and aggressive glide path.The moderate lifecycle path consists of the same asset allocation path analyzed in Figure 1. The equity exposure of all three lifecycle paths is shown in Figure 5. Figure 6 shows the income replacement ratio at the age of 65 of each of the glide paths.
䉴 FIGURE 5
EQUITY ALLOCATION OF LIFECYCLE PATHS
The expected replacement ratio ranges from 5265%, while the probability of meeting the 60% replacement target ranges from 40-55%. As expected, equity variation among lifecycle paths results in wide dispersion of replacement ratios. Greater dispersion in replacement rates is observed on the upside (25th and 5th percentiles) rather than the downside. Over a long time horizon with the compounding of returns, a more aggressive lifecycle path has significantly more upside potential. On the downside (95th percentile), the worst-case replacement ratio for the aggressive path is no worse than the conservative and moderate paths. The aggressive path may be more volatile than the other paths, but the long time horizon leads to greater asset accumulation. The aggressive path can achieve a replacement ratio similar to that of the conservative path in the worst-case scenario simply because it has more assets to lose when disaster strikes. An aggressive lifecycle path is attractive from a long-term perspective but would have high short-term volatility during the eligible retirement period. Individuals nearing retirement become more sensitive to short-term volatility and may prefer a more conserva-
tive lifecycle path from ages 50 to 65. Figure 7 shows the one-year percentage decline in assets in a worstcase scenario.7 A potential one-year decline (5% chance) in assets of 9-12% from ages 50 to 65 may be too high for some investors. Plan sponsors may wish to take shortterm volatility into consideration as they evaluate the appropriateness of glide paths. Plan sponsors may find that younger participants are also sensitive to short-term volatility. Research on myopic risk aversion suggests that a long-term investment horizon associated with retirement does not necessarily mean that participants will be insensitive to near-term losses.8 Plan sponsors that are concerned about their participants’ ability to stay the course during market downturns may wish to weigh interim volatility more heavily than those that believe their DC participants will have a long-term focus. Plan sponsors may wish to consider employee turnover when selecting the aggressiveness of the glide path. According to a Hewitt Associates study, 45% of 401(k) participants take a cash distribution upon cessation of employment (this number is even higher for younger workers).9 If cashouts are likely,
BENEFITS QUARTERLY, Third Quarter 2007
39
䉴 FIGURE 6
INCOME REPLACEMENT RATIO AT AGE 65
Percentile
Conservative Path
5th (Best Case) 111% 25th 71% 50th (Expected Case) 52% 75th 39% 95th (Worse Case) 26% Probability 40% (replacement ratio ⬎ 60%)
short-term volatility may figure more prominently in the glide path selection equation. Fortunately, employees have many options for preserving the tax-deferred status of their retirement assets upon cessation of employment. Former employees may be able to leave assets in their old 401(k) plan or roll over to another tax-deferred retirement vehicle, such as the successor employer defined contribution plan or an individual retirement account (IRA). A lifecycle path should be developed with the primary goal of retirement income adequacy. Also note, in Figure 7 the lifecycle paths are projected to experience smaller and smaller possible one-year declines in assets over time, as their equity allocation declines. Because the balanced fund maintains a constant level of equity exposure, this leads to a perverse increase in risk as retirement approaches. The potential for a larger one-year decline in assets under the balanced fund is a result of assets trending
40
BENEFITS QUARTERLY, Third Quarter 2007
Moderate Path
Aggressive Path
143% 86% 60% 43% 26% 50%
171% 97% 65% 44% 26% 55%
upward under a constant level of equity exposure. The balanced fund may be too aggressive for participants over the age of 60.
OTHER ASSET CLASSES AND DIVERSIFICATION Another possible way to improve retirement adequacy is through diversification. So far, our lifecycle paths have only consisted of two asset classes. Figure 8 introduces six new asset classes to the moderate lifecycle path. In Figure 8, the moderate path’s equity exposure is diversified into small/mid cap U.S. equity and nonU.S. equity. Not only does this improve overall diversification, the potential for adding excess return, or return above the benchmark, in small/mid cap U.S. equity and non-U.S. equity is higher than for large cap U.S. equity. In the context of the overall portfolio, other ef-
䉴 FIGURE 7
ONE-YEAR PERCENT DECLINE IN ASSETS WORST-CASE SCENARIO (95TH PERCENTILE)
fective diversifiers include real estate investment trusts (REITs), treasury inflation-protected securities (TIPS) and high-yield bonds. As hybrid securities, REITs and high-yield bonds have risk/return characteristics that are between equity and U.S. fixed income and can enhance or maintain asset growth if funded from a combination of equities and U.S. fixed income. The allocation to REITs and the ratio of high-yield bonds to U.S. fixed income is reduced from age 50 onward to dampen portfolio volatility. TIPS are not a growth asset but rather serve as an effective hedge against both expected and unanticipated inflation. Equities are an effective longterm hedge against inflation, but the equity exposure in the glide path begins to decline dramatically from age 50 onward. A participant is most vulnerable to inflation from ages 50 to 70, when the participant has built a large account balance due to years of prefunding and investing. TIPS are also a lowrisk asset (low volatility and weak correlations to other asset classes) and can be used to reduce glide
path volatility. The value of TIPS to older participants results in an allocation that emerges at the age of 50 and gradually grows as a percent of the portfolio thereafter. Short-duration bonds are introduced after the age of 50 to further dampen portfolio volatility. The allocation to short-duration bonds increases as retirement approaches and the participant’s time horizon (life span) shortens. Figure 9 shows the range of income replacement ratios for the moderate-diversified path compared to that of the nondiversified moderate and aggressive paths. The moderate-diversified path is expected to replace 7% and 12% more income than the nondiversified aggressive and moderate paths, respectively. The probability of achieving the 60% replacement target increases 7% to 12% by moving to the moderatediversified path as well. Most impressive of all, the moderate-diversified path has significantly more upside potential (5th percentile) and a higher replacement ratio in the worst-case scenario. Clearly,
BENEFITS QUARTERLY, Third Quarter 2007
41
䉴 FIGURE 8
MODERATE-DIVERSIFIED LIFECYCLE PATH
diversification and thoughtful asset allocation can significantly improve a lifecycle path.
ALTERNATIVE INVESTMENTS As lifecycle paths begin to push the envelope as far as employing alternative investments within the DC plan structure, they face challenges such as the need for daily valuation and daily liquidity. For example, private equity programs involve a 15-20 year time horizon with capital calls and net distributions at periods set by the manager, not the investor. Likewise, hedge funds typically have lockup periods and limited opportunities for contributions and redemptions. Finally, private real estate is subject to infrequent appraisal smoothing. Currently, a great deal of energy and time is being expended by product innovators and other interested parties to incorporate illiquid asset classes (or a close representation) into a DC plan structure. In the near future, it may be possible to include traditionally illiquid asset classes in the lifecycle path. Plan sponsors, especially those that are building lifecycle paths out of their underlying core investment options, should stay alert to any developments.
42
BENEFITS QUARTERLY, Third Quarter 2007
POSTRETIREMENT ASSET ALLOCATION The event of retirement is an important inflection point in the glide path, as the investor shifts from an accumulation phase to a spending phase. At this stage of the lifecycle, the risk of outliving assets (longevity risk) and the erosion of purchasing power due to inflation are primary threats to future financial security. A significant allocation to assets with capital growth potential may thwart these risks. At the same time, the ability to assume market risk is more limited for a retiree. Unlike the situation of the young investor with an ample recovery period and assets flowing into the 401(k) plan, a retiree is spending down the assets. Meanwhile, a contracting life expectancy leads one to question the sufficiency of the recovery period. A severe market downturn that coincides with spending down assets can deplete a participant’s balance to such a low level that an older retiree may not have sufficient time to recoup investment losses. Figure 10 shows the probability of depleting the participant’s balance to zero (starting at age 65 retirement) by ages 70 to 105 years for four lifecycle paths
䉴 FIGURE 9
INCOME REPLACEMENT RATIO AT AGE 65
Percentile
Moderate Path
5th (Best Case) 143% 25th 86% 50th (Expected Case) 60% 75th 43% 95th (Worse Case) 26% Probability 50% (replacement ratio ⬎ 60%) examined earlier. In this analysis, the participant is assumed to spend 60% of final salary each year directly from the 401(k) plan. In addition, spending is assumed to increase each year with simulated inflation to keep pace with a rising cost of living. As expected, the probability of depleting assets to zero rises as a participant gets older. The life expectancy of someone aged 65 is 18 years (the age of 83).10 In other words, 50% of participants are expected to live beyond the age of 83. The moderatediversified path stands out, since it has the lowest probability of asset depletion and with less short-term volatility than the aggressive path. Interestingly, the conservative path is the most risky, as the probability of asset depletion is the highest. This is in stark contrast to Figure 7 where the conservative path was the low-risk option during ages 50 to 65. In Figure 7, risk was defined as a oneyear decline in assets (a short-term measure of risk). The cost of lower short-term volatility is lower asset growth prior to the age of 65, resulting in a higher
Aggressive Path
ModerateDiversified Path
171% 97% 65% 44% 26% 55%
197% 105% 72% 49% 29% 62%
probability of depleting assets prematurely. Although low volatility may be desirable as retirement approaches, investors have to keep in mind that they may live another 20-30 years beyond the age of 65.
SAVINGS RATES The analysis thus far has assumed contribution levels of 9% of pay (6% employee and 3% employer contributions) until the age of 65. In this section, the authors contrast the retirement income adequacy achieved with a 5%, 9% and 13% savings policy. The moderate-diversified (“Mod-Div”) lifecycle path is the underlying default investment option in each case. Once again, the range of replacement rates is shown in Figure 11. The level of savings clearly has a dramatic impact on retirement adequacy. The probability of achieving a 60% replacement rate increases from 23% to 62% when contributions are increased from 5% of pay to 9% of pay. A further increase of 4% (from 9% of pay
BENEFITS QUARTERLY, Third Quarter 2007
43
䉴 FIGURE 10
PROBABILITY OF ASSET DEPLETION
to 13% of pay) increases the probability of hitting the 60% target from 62% to 83%. Increasing contributions also shifts the entire range of probable replacement ratios upward leading to less downside risk (higher replacement in the 95th percentile) and significantly more upside potential. Figure 11 is an important reminder that a robust default contribution rate and the use of automatic contribution escalation are critical factors to consider when developing automatic enrollment programs.
CONCLUSION As the U.S. retirement system moves from a defined benefit to a defined contribution focus, the authors find a simultaneous evolution of defined contribution plans incorporating many of the features traditionally associated with defined benefit plans. Automatic pension plan participation is being mirrored by automatic enrollment in the DC environment. Payouts that increase with pay in the DB world are mirrored with automatic contribution escalation in the DC environment.
44
BENEFITS QUARTERLY, Third Quarter 2007
By using retirement income replacement analysis as a means of evaluating default investment options, plan sponsors can take a page from the asset/liability type analysis that has long been a cornerstone of pension plan management. Just as asset/liability analysis seeks to match an appropriate asset allocation with pension funding needs, retirement income replacement analysis seeks to provide valuable insight into how well a series of lifecycle funds will meet defined contribution plan participants’ retirement funding needs. Of course, there are additional factors that need to be addressed in selecting an appropriate default option: the risk tolerance and demographics of the employee population, the degree of employer paternalism and the existence of other retirement vehicles to support and enhance retirement adequacy. One thing is clear, however: The days of assuming that the impact on retirement income adequacy is negligible when it comes to the default option are long gone. Going forward, the selection of the default investment option will be one of the most critical decisions that plan sponsors make when designing their defined contribution plan. 䉳
䉴 FIGURE 11
INCOME REPLACEMENT RATIO AT AGE 65
Percentile
ModerateDiversified (5% of Pay)
5th (Best Case) 110% 25th 59% 50th (Expected Case) 40% 75th 27% 95th (Worse Case) 16% Probability 23% (replacement ratio ⬎ 60%)
Endnotes 1. Callan’s proprietary Monte Carlo simulation model is employed to simulate investment returns and inflation across 1,000 economic scenarios. 2. To determine income replacement ratios, the account balance must first be converted to a life income stream. Annuity assumptions are a 5% interest rate, RP-2000 mortality with 50/50 blend (male/female) and 2.75% contractual indexing. (Inflation is expected to be 2.75% per year.) 3. Assumed interest rate for annuity conversion is 5% across all simulated scenarios. In a higher (lower) interest rate environment, annuities are cheaper (more expensive) and lead to higher (lower) replacement rates. 4. Hewitt Associates’ 2005 Research Report: Total Retirement Income at Large Companies: The Real Deal.
ModerateDiversified (9% of Pay)
ModerateDiversified (13% of Pay)
197% 105% 72% 49% 29% 62%
285% 152% 104% 71% 41% 83%
5. Measuring Retirement Income Adequacy: Calculating Realistic Income Replacement Rates. Jack VanDerhei. EBRI Issue Brief, No. 297, September 2006. 6. The Employee Benefit Research Institute’s 2006 Retirement Confidence Survey. 7. A linear line is fitted to each year’s simulated worst-case decline in assets by least squares regression. 8. “The Effect of Myopia and Loss Aversion on Risk Taking: An Experimental Test.” Richard Thaler, Amos Tversky, Daniel Kahneman and Alan Schwartz. Quarterly Journal of Economics. Volume 112, pages: 647-61. 9. “Are Workers Preserving Their 401(k) Retirement Wealth?” Hewitt Associates. July 2005. 10. Based on RP-2000 Mortality table published by the Society of Actuaries. Mortality rates are blended 50/50 male/female.
International Society of Certified Employee Benefit Specialists Reproduced from the Third Quarter 2007 issue of BENEFITS QUARTERLY, published by the International Society of Certified Employee Benefit Specialists. With the exception of official Society announcements, the opinions given in articles are those of the authors. The International Society of Certified Employee Benefit Specialists disclaims responsibility for views expressed and statements made in articles published. No further transmission or electronic distribution of this material is permitted without permission. Subscription information can be found at www.iscebs.org. ©2007 International Society of Certified Employee Benefit Specialists BENEFITS QUARTERLY, Third Quarter 2007
45
