Revisiting the To vs. Through Debate: Our Approach to the Target-Date Glide Path

Transcription

1 JUNE 2013 Revisiting the To vs. Through Debate: Our Approach to the Target-Date Glide Path Rich Weiss Senior Vice President Senior Portfolio Manager Asset Allocation Strategies Nancy Pilotte Vice President Client Portfolio Manager This paper is an update and expansion of our earlier piece titled Our Approach to the Target-Date Glide Path. The original paper addressed the difference in to versus through retirement glide paths in proprietary target-date funds (TDFs). Further, the paper laid out the rationale for our approach to the construction of the glide paths of the One Choice SM Target Date Portfolios. Our philosophy and approach have not changed. What is new is that the intervening years have provided additional academic research on the slope of the retirement glide path and its relation to retirement outcomes. We ve updated the original paper to reflect that new research. We ve also incorporated some of our own recent work on the path dependency of market returns for investor outcomes, and addressed how this relates to glide path slope. Our analysis and the latest third-party research indicate that the best outcomes for retirees in target-date portfolios are achieved when the portfolio reaches its most conservative asset allocation at retirement, as opposed to entering retirement with a risky allocation and gradually decreasing equity exposure throughout the decumulation phase. Furthermore, it is clear that a static asset allocation with fairly balanced equity and fixed-income weightings postretirement increases the likelihood of a fully-funded retirement. Other things equal, the flatter glide path reduces the volatility or dispersion of retirement wealth outcomes for plan participants. Simply put, the evidence runs in favor of the to retirement glide path, which connotes more conservative equity exposure in the years leading up to the target date, and a flat glide path in retirement. By comparison, a through retirement glide path tends to be too aggressive in the years just before and just after retirement, and too conservative later in retirement.

2 Defining Terms Target date: The target date or year within the fund name is the approximate date when the typical investor in the fund plans to start withdrawing money. This does not mean that the fund cannot be held past the retirement date. Rather, the typical target-date fund (TDF) is designed so that investors may continue in the portfolio and begin to draw down their balance gradually over time to fund their retirement needs. To glide path: TDFs in the to camp are managed with a glide path that reaches its most conservative asset allocation at the funds target date, and remains at a fixed allocation thereafter. It is important to understand that most TDFs with to glide paths are designed to fulfill post-retirement investment and income needs and are continually managed and monitored throughout the entire lifecycle. In many cases, this is achieved by folding the target-date fund into a retirement income portfolio within a few years after the target-date is reached. Through glide path: TDFs in the through camp do not reach their most conservative asset allocation until after the target date. Relative to TDFs in the to camp, they tend to hold more equities in the five to 10 years before and after the maturity date, and to hold fewer equities in the later retirement years. Age Versus Wealth a Determinant of Risk Exposure Age level is merely a proxy for one of the true drivers of risk tolerance wealth level. It is well established in theory and practice that as wealth levels increase (and retirement nears), investors rightfully become more risk averse. It is this increased aversion to risk that drives the asset allocation decision towards less risky investments. As one s age progresses, and accumulated wealth decreases post-retirement, the relationship between age and risk aversion in not as strong. (See, for example, Wang and Hanna, 1997; Hariharan, Chapman, and Domian, 2000; and Khanapure 2012.) Through Glide Path Emphasizes Age Over Wealth Proponents of the through methodology are keenly concerned with longevity risk. That is, the risk that investors will run out of money in retirement. They argue that longer lifespans in retirement and rapidly rising health care costs justify higher equity allocations in the 10 to 15 years around the target date, while risk aversion and the need for income-producing investments mean higher bond allocations later in retirement. These practitioners are more concerned with an investor s age than wealth level. It can be argued that the through approach harkens back to the old rule of thumb for lifecycle investing: Equity allocation should equal 100 minus your age. The notion that equity allocation (and portfolio risk in general) should be reduced as one ages is certainly not new, but modern investment theory has advanced well beyond that antiquated concept which naïvely and directly ties asset allocation to age. To Approach Sensitive to Wealth Level The to approach, on the other hand, recognizes the key role that wealth level (not age, per se) plays in asset allocation and, therefore, reduces risk inversely to wealth accumulation. At the peak wealth level (defined as retirement date in most scenarios), the to approach appropriately stands at its least risky posture in terms of asset allocation. While we and other members of the to camp agree that stocks can play a significant role in growing capital and preserving purchasing power throughout retirement, the allocation decision must take into account the risk tolerance of investors nearing or just past retirement. Indeed, we find that investors are most vulnerable at retirement, and therefore, high equity exposure introduces unwanted market risk. As a result, we reject the notion that time is still on your side at or near retirement and that high equity allocations are appropriate in the critical years just before retirement. Instead, we believe a to retirement allocation takes a more balanced approach to resolving the tension between longevity and market risks. To see why this is so, we consider the effect of market volatility on a hypothetical portfolio in the run-up to retirement, and in the immediate aftermath of retirement, concluding that TDFs should reach their most conservative allocation at retirement. A Conservative Allocation Is Critical at Retirement Market risk becomes most critical as wealth levels increase and contributions cease. Investors typically accumulate 50% of their retirement wealth in the last seven years prior to retirement. As a result, the future ability to fund a long retirement is more affected by a significant shock to the portfolio in this period than at any other point along the glide path. This is so because from a purely financial perspective, the point of greatest peril is the day you retire: Your balance is at its highest level, and you have the 2

3 longest period to fund in retirement. Significant losses at this juncture can effectively erase years of retirement income. In addition, a shock in this period is much more difficult to recover from since you will start withdrawing from, rather than contributing to, your account. As a result, a market shock early in retirement can cause significantly more damage to the longevity of the portfolio than a shock later on. In Figure 1, we see that a 15% downside shock at age 67 shortens the life of the portfolio by five years, depleting the balance at age 85 instead of it lasting past age 90. By comparison, the same 15% shock at age 80 has less of an impact, shortening the life of the portfolio by two years. Again, higher sensitivity to a loss occurs early on in retirement when wealth levels are greater. Figure 1: Conservative Allocation Crucial to Managing Risk Early in Retirement $1,000,000 $800,000 $600,000 Studies in behavioral finance confirm that investors are prone to bailing out of portfolios that have incurred one or two years of losses. Cumulative Balance $400,000 $200,000 $0 -$200,000 -$400,000 -$600,000 -$800, Age Base Case -$62,418 Shock at 80 -$275,599 Shock at 67 -$631,196 Source: American Century Investments Base case scenario assumptions: Assumes annual assets return of 5%, first year withdrawal of 4% of capital, increased by 3% annually. Shock: A single-year -15% event. This information is for illustrative purposes only and is not intended to represent any particular investment product. Behavioral Argument for To Glide Path What s more, investors are generally loss averse. This is particularly true when losses are large in dollar terms, and investors have a sense that they have little time to recover from those losses as is true in the years just prior to and after retirement. Studies in behavioral finance confirm that investors are prone to bailing out of portfolios that have incurred one or two years of losses. In addition, behavioral studies have shown that investors who check their balances frequently tend to reduce their allocations to equities, while investors who check their balances infrequently tend to allocate more to equities. This is known as myopic loss aversion, and could help explain why the category of target-date funds experienced sizeable net redemptions in the 2008 market slide, according to Morningstar, with many investors locking in losses that averaged -22% for 2008 (Thaler, Tversky, Kahneman, and Schwartz; The Effect of Myopia and Loss Aversion on Risk Taking: An Experimental Test ; Quarterly Journal of Economics; Vol. 112, Issue 2, 1997, and Morningstar Fund Analysts; Target-Date Investors Stick Around, Earn Better Returns ; March 16, 2010). For these reasons, in constructing the One Choice Target Date Portfolios, we have purposely avoided the overly aggressive equity allocation investment approach built into TDFs in the through camp. Rather, we conclude that target-date funds should reach their most conservative allocation at retirement, at the target date. 3

4 Path Dependency Also Supports Flatter Glide Path In this section, we look further at how a flatter glide path helps minimize variations in retirement wealth caused by the sequence of investment returns. Not only is a flatter glide path preferable post-retirement, but our analysis shows that a relatively flatter glide path during the accumulation phase (pre-retirement date) helps to insulate participants from sequence of returns risk. Otherwise known as path dependency, this risk is a major source of volatility in target-date fund investing. The exact order in which returns are experienced can significantly impact an investor s wealth accumulation. Figure 2 shows how significant an impact the pattern of market returns can have on investor account balances, even in the years before retirement. The graph depicts investment outcomes using actual and reverse chronological S&P 500 Index returns (and therefore having the same arithmetic average return) with regular contributions over the period. $3,500,000 Figure 2: Pattern of Return Matters Greatly Actual w $10,000/Yr Reverse w $10,000/Yr Figure 2: This graphic shows just how significant an impact the pattern of market returns can have on investor account balances, even in the years before retirement. This chart depicts investment outcomes using actual and reverse chronological S&P 500 returns (and therefore having the same arithmetic average return) with regular contributions over the period. $3,000,000 $2,500,000 $2,000,000 $1,500,000 $1,000,000 $500,000 $0 Dec-82 Dec-83 Dec-84 Dec-85 Dec-86 Dec-87 Dec-88 Dec-89 Dec-90 Dec-91 Dec-92 Dec-93 Dec-94 Dec-95 Dec-96 Dec-97 Dec-98 Dec-99 Dec-00 Dec-01 Dec-02 Dec-03 Dec-04 Dec-05 Dec-06 Dec-07 Dec-08 Dec-09 Dec-10 Dec-11 Dec-12 Hypothetical illustration. Source: American Century Investments. The chart depicts the growth of $10,000 invested in the S&P 500 annually from 12/31/1982 to 12/31/2012. Actual refers to the actual returns of the S&P 500. Reverse depicts the annual returns of the S&P 500 in reverse chronological order. This information is for illustrative purposes only and is not intended to represent any particular investment product. There can be no assurance similar results can be achieved, and this information should not be relied upon as a specific recommendation to buy or sell an investment product. Next, consider a hypothetical investor who holds a constant 50/50 stock/bond mix throughout her lifetime. In effect, the investor is relatively indifferent to the exact sequence of equity and bond returns since she is always, by definition, equally weighted. (This investor s portfolio corresponds to the left endpoint of the line on Figure 3.) On the other hand, imagine an investor who begins with an all-equity portfolio and systematically reduces that equity allocation over time, eventually ending with 0% equity and 100% fixed income (corresponding to the right most point on the line graphed in Figure 3). In that case, the investor clearly prefers (i.e., needs) good equity returns early on, when he holds a high allocation to that asset class. Poorer equity returns later in their life are much less relevant, given his reduced exposure. The exact sequence, or path, of equity returns will play a more influential role in the latter portfolio than in the former. 4

5 The relationship of glide path slope to variation in wealth outcomes is plotted in Figure 3. Using Monte Carlo simulation techniques, we examined the degree to which slope over the life of the glide path is related to ending wealth variation. Beginning on the left side of the graph, we see that a perfectly flat glide path (i.e., one that remains equally balanced between stocks and bonds throughout the lifecycle) can experience a 14% variation in ending outcomes, depending on the exact sequence of returns assumed. As one steepens the glide path slope (moving up and to the right on the line plotted), the variation in ending wealth outcomes increases significantly. In fact, steeper glide paths introduced up to 75% more uncertainty in outcomes at retirement. 30% Figure 3: Effect of Glide Path Slope on Retirement Outcomes Rolling 30-Year Historical Returns, Figure 3: Flatter glide paths provide more certainty around the outcome. Steeper glide paths introduced up to 75% more uncertainty in outcomes at retirement. Volatility of Retirement Wealth Outcome 25% 20% 15% 10% 5% Flat Steep Glide Path Slope Source: American Century Investments. Glide paths are divided between stocks (S&P 500 Index) and bonds (Intermediate Government). Several glide paths are examined, and are denoted by their beginning (year 0) and ending (year 30) stock allocation. For example, the glide path is flat; it begins and ends with 50% stock allocation. The glide path is steeply sloped, beginning with 100% stock allocation and ending with no stock allocation. Volatility of retirement wealth outcome for each set measured as the standard deviation of year-to-year change in 30-year cumulative return. Returns are simply compounded over 30-year rolling windows. This test was done for the overlapping 30-year windows from The projections generated by American Century Investments Monte Carlo simulation regarding the likelihood of various investment glide paths are hypothetical in nature, do not reflect actual results, and are no guarantee of future results. Glide Paths in Retirement: Flat Better than Sloping Finally, we discuss the effect of glide path slope in the post-retirement period. Our research indicates that a flat glide path in retirement is superior to a continually sloping glide path. We conducted Monte Carlo simulations of expected outcomes for a hypothetical portfolio beginning with a balance of $700, years prior to the target date. We assumed a contribution of $15,000 per year for the remaining 15 years, and a withdrawal of $50,000 per year for 20 or 30 years after the target date (all dollars in real terms). We constructed various equity/bond allocation glide paths that continued to reduce equity exposure through retirement, and one that became static after retirement, like the One Choice Target Date Portfolios glide path, shown in Figure 4. By keeping the average equity exposure roughly equal in the various to and through glide paths, the median expected outcomes over the 35- and 45-year simulation horizons were all statistically similar. However, we found that in every case, the flat One Choice glide path in retirement resulted in a lower percentage of zero-balance outcomes. 5

6 In the sloping glide path scenarios, such as the one shown in Figure 4, the likelihood of running out of money in retirement was 10%, or twice that of our flat glide path scenario, which had a 5% probability of depleting the portfolio. This crucial point bears repeating: Even though the average equity exposure over the life of the portfolio was the same in the flat and sloping glide paths we analyzed, the flat glide path reduced by half the likelihood that a retiree will run out of money soon. A 2010 research paper from Russell Investments (Cohen, Gardner, and Fan; The Date Debate; April 2010) approaches the to versus through debate in slightly different terms, asking instead What is the investment rationale for a sloping versus flat postretirement glide path?, and How aggressive should post-retirement allocations be? To answer these questions, the authors examine a number of scenarios analyzing the performance of different portfolios with flat and downward-sloping risk profiles across varying time horizons and withdrawal assumptions. Their findings are remarkably consistent across these scenarios, and dovetail with our own proprietary research. Specifically, the study finds that a flat glide path in retirement always makes sense relative to a sloping one without regard to the level of aggressiveness. In fact, the authors state, there is not a clear investment rationale for the glide path to slope. Moreover, they conclude that for each downward-sloping glide path there is a corresponding flat glide path that gives a higher expected ending wealth for the same amount of risk. Also, there is a flat glide path that provides the same expected ending wealth for a lower level of risk. Figure 4: Typical Contrasting Glide Paths of To and Through Portfolios Over Time Prob. of $0 outcome Summary Stats: ($ million) One Choice Target Date Portfolios Through Retirement One Choice Target Date Portfolios Through 5% 10% Max $16.5 $21.2 Median $1.5 $1.5 Stdev $2.1 $2.3 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Years (Pre)/Post-Retirement Source: American Century Investments Equity Allocation Another important aspect to favoring a flat glide path in retirement is the fact that financial situations grow increasingly diverse in retirement. Because retirement plan fiduciaries and pension managers whose success or failure is ultimately defined by providing the best outcomes for retirees cannot know the specific financial condition of each account holder, a flat, balanced allocation is the safest approach. This sort of allocation and glide path is well suited to form a core holding in a customized retirement 6

7 plan devised by account holders or their financial advisor to best meet each person s specific needs. Contrast this with the constantly changing allocation of a TDF managed through retirement and imagine the retirement planning complexities this can introduce. Conclusions Our own research and a growing body of evidence suggest that target-date portfolios managed to retirement rather than through retirement generate better outcomes for retirees; that is, to TDFs tend to maximize the likelihood of retirement success for plan participants without taking undue risk along the way. This reflects the fact that to TDFs typically take less risk in the crucial years immediately before and immediately after retirement. Additionally, the flatter glide path trajectory in the One Choice Target Date Portfolios, both pre- and post-retirement date, produces lower volatility along with a greater likelihood that retiree account balances will last throughout retirement. Material presented has been derived from industry sources considered to be reliable, but their accuracy and completeness cannot be guaranteed. Past performance is no guarantee of future results. This information is not intended to serve as investment advice. You should consider the fund s investment objectives, risks, and charges and expenses carefully before you invest. The fund s prospectus or summary prospectus, which can be obtained by visiting americancentury.com/ipro, contains this and other information about the fund, and should be read carefully before investing. The performance of the portfolios is dependent on the performance of their underlying American Century Investments funds and will assume the risks associated with these funds. The risks will vary according to each portfolio s asset allocation, and a fund with a later target date is expected to be more volatile than one with an earlier target date. The target date is the approximate year when investors plan to start withdrawing their money. The principal value of the investment is not guaranteed at any time, including at the target date. Each portfolio seeks the highest total return consistent with its asset mix. Each year, the asset mix and weightings are adjusted to be more conservative. In general, as the target year approaches, the portfolio s allocation becomes more conservative by decreasing the allocation to stocks and increasing the allocation to bonds and cash alternatives. 7

8 Selected Bibliography Basu, Anup K. and Michael E. Drew, Portfolio Size Effect in Retirement Accounts: What Does It Imply for Lifecycle Asset Allocation Funds? The Journal of Portfolio Management, Spring 2009, pp Bodie, Zvi and Jonathan Treussard, Making Investment Choices as Simple as Possible, but Not Simpler, Financial Analysts Journal 63, 2007, pp Bodie, Zvi, Richard K. Fullmer and Jonathan Treussard, Unsafe at Any Speed? The Designed-In Risks of Target-Date Glide Paths, Journal of Financial Planning, March 2010, pp Cohen, Josh, Grant Gardner and Yuan-An Fan, The Date Debate, Russell Research, April Ellement, Jason L. and Lori Lucas, The Great Target-Date Fund Debate, Benefits Quarterly, Fourth Quarter 2009, pp Hariharan, Govind, Chapman Kenneth S., and Domian, Dale L., Risk Tolerance and asset allocation for investor nearing retirement, Financial Services Review 9, 2000, pp Idzorek, Tom, Target Date Solutions: Is a Target Date Enough? CFA Institute Conference Proceedings Quarterly, September 2009, pp Jacobsen, Brian, Christian Chan, and Olivia Barbee, Balancing Longevity Risk and Market Risk: The Impact of the PPA on Target Date Funds, Journal of Pension Benefits, 2009, pp Khanapure, Revansiddha, What Drives Risky Investments Lower at Retirement?, August 31, Available at SSRN: or Lewis, Nigel D., Making ends meet: Target date investment funds and retirement wealth creation, Pensions, Vol. 13, 2008, pp Pang, Gaobo and Warshawsky, Mark J., Asset Allocations and Risk-Return Tradeoffs of Target- Date Funds, October 15, Available at SSRN: Poterba, James, Joshua Rauh, Steven Venti and David Wise, Lifecycle Asset Allocation Strategies and the Distribution of 401(k) Retirement Wealth, NBER Working Paper No , January Spitzer, John J. and Sandeep Singh, Shortfall risk of target-date funds during retirement, Financial Services Review 17, 2008, pp Thaler, Richard H., Amos Tversky, Daniel Kahneman, and Alan Schwartz; The Effect of Myopia and Loss Aversion on Risk Taking: An Experimental Test, Quarterly Journal of Economics, Vol. 112, Issue 2, Viceira, Luis M., Life-Cycle Funds, Harvard Business School, NBER and CEPR, May Wang, Hui and Hanna, Sherman, Does Risk Tolerance Decrease with Age?, Financial Counseling and Planning, Volume 8(2), Yoon, Youngjun, Glide Path and Dynamic Asset Allocation of Target Date Funds, October 28, Available at SSRN: Behavior. Opinions expressed are those of Rich Weiss and Nancy Pilotte and are no guarantee of the future performance of any American Century Investments portfolio. Opinions and estimates offered constitute our judgment and, along with other portfolio data, are subject to change without notice. American Century Investment Services, Inc., Distributor 2013 American Century Proprietary Holdings, Inc. All rights reserved. IN-WHP