CHAPTER 2
Aligning DC Investment Design to Meet the PRICE of Retirement

Solving every asset allocation problem is the same mutatis mutandis—that is, changing only those things that need to be changed. There is always an objective for the investment, which is either real or nominal. For DC participants, the investment objective is usually to replace income in retirement—in real terms. Risk is the failure to meet the objective. Then DC investment choices, as well as the asset allocation and constraints, can be defined to align to the objective and within the risk capacity of the investors.

—Harry Markowitz, Nobel Prize–winning economist and father of modern portfolio theory

In 2013, I had the great honor, along with colleagues Ying Gao and Michael Esselman, of meeting Nobel Laureate and acknowledged father of modern portfolio theory Professor Harry Markowitz in his San Diego office. As plan fiduciaries consider the objective for their DC plan and structure the plan’s investments to help meet it, Markowitz’s words are influential. He tells us that for DC participants, the objective is to replace income in retirement—in real terms. In other words, a DC plan investment lineup should be designed and managed to meet a retirement income objective—one that builds sufficient assets to maintain a participant’s lifestyle during retirement. To accomplish this goal, DC assets and the retirement income distributions must keep pace with inflation.

Image described by the caption.

FIGURE 2.1 My 2013 Meeting with Dr. Harry Markowitz

Left to right: Michael Esselman, Dr. Harry Markowitz, Stacy Schaus, and Ying Gao (PIMCO).

Professor Zvi Bodie, The Norman and Adele Barron Professor of Management at Boston University (now retired), also underscores the importance of inflation protection in DC plans. In fact, he notes that DC plans offer an advantage relative to most defined benefit plans that lack inflation protection. Bodie shared with us (in PIMCO’s March 2007 DC Dialogue): “Inflation protection makes the DC option stronger than traditional DB plans in this country, because DB plans don’t offer inflation protection typically. Rather, a retiree’s retirement income erodes over time as his or her dollars buy less in retirement.”

Professor Bodie goes on to recommend that DC plans offer, as a retirement risk-free asset, treasury inflation-protected securities (TIPS) as these individual bonds contractually keep pace with inflation. He identifies TIPS as the most prudent asset for retirement investing. In Bodie’s book, Worry-Free Investing, he tells readers that “If you want to sleep nights secure in the knowledge that you will achieve your savings goal, you must invest in a way that eliminates the possibility that inflation will undercut your efforts. If you try to do it by saving less and expecting the stock market to do the heavy lifting, you may not get there at all.” He adds, “There are circumstances when you will want to take calculated risks by investing in the stock market in the hope of increasing your future income or wealth. . . . . But if you want to hit specific targets for certain, then worry-free investing with inflation-protected securities is the way to go.”

Both Professors Markowitz and Bodie influence and guide our thinking about investing for retirement, starting with the objective to create a real income stream in retirement and seeking appropriate assets to help meet this objective.

In this chapter, we’ll take a close look at the DC plan objective and setting a specific target replacement income rate. Then we’ll suggest a methodology to quantify the income-replacement target using the PIMCO Retirement Income Cost Estimate (PRICE). We will show you how the PRICE methodology can help determine how much it may cost to retire, whether the investment default and other plan investments are aligned to meet a retirement income objective, and finally how to consider whether the retirement income offerings are likely to keep pace with the cost of retirement.

BEGIN WITH THE END IN MIND

As Steven Covey tells us in The 7 Habits of Highly Effective People, we should “begin with the end in mind.” We agree with this thinking and encourage plan fiduciaries to think first about the investment objective for their DC plan. As discussed earlier, for most that objective is to help provide sustainable retirement income for participants. Over the past 10 years, we have interviewed dozens of DC plan sponsors and consultants about plan design. In the DC Dialogue series, we often begin with the question “What is the philosophy for design or objective for your DC plan?” In PIMCO’s November/December 2012 DC Dialogue, Karin Brodbeck, Director of Retirement Investments at Nestlé USA, Inc., shared Nestlé’s philosophy for retirement plan design:

Our management style is to look at our retirement plans holistically. For most workers, we have a three-legged retirement stool, including Social Security, DB, and DC. Speaking specifically to DC, our general philosophy is that our plans need to be designed to deliver a reasonable likelihood of meeting the participants’ income needs. As we designed our plans, we aimed for the combined sources of income to replace 70 percent of an employee’s final pay. For those who only have a DC plan, they may need 40 percent of their pay replaced by the DC plan.

While many plan sponsors consider the DC plan the primary retirement plan for employees, they often do not state a specific income replacement target. However, they may be aware of the retirement income replacement need and take this into account as they design their plan and investment defaults. For instance, Brad Leak, CFA, managing director of public markets at The Boeing Company, tells PIMCO in a 2012 DC Dialogue that “We don’t have a specific income replacement goal. However, we did consider whether various glide paths were likely to maintain a retiree’s consumption pattern throughout retirement. We focused on glide path designs that had at least a 90 percent chance of maintaining the modeled consumption pattern.” (The term “glide path” refers to a formula that defines the asset allocation mix of a fund, based on the number of years to a target date. The glide path typically creates an asset allocation that becomes more conservative—that is, includes more fixed-income assets and fewer equities—the closer a fund gets to the target date.)

Generally, corporate plan sponsors tell us that the defined benefit plan is no longer open to new hires and instead participants must look to the DC plan as the primary occupational pension scheme. But what about employers that do offer a DB plan? Is their DC plan objective different? Not necessarily.

Over the years, we’ve spoken with many DC plan sponsors that also offer a DB pension plan to participants. For instance, nearly all of the public employers continue to offer a DB benefit. Nonetheless, these plan sponsors acknowledge that retirees increasingly will look to DC plans to replace at least a certain percentage of a participant’s final pay. In the May/June 2015 DC Dialogue, Joshua Franzel, PhD, Vice President of Research at the Center for State and Local Government Excellence (slge.org) in Washington, DC, shared an overview on retirement income sources for public workers:

Public workers may have one or more sources of retirement income, including Social Security, a DB plan, a DC plan, or a hybrid program, which may be a DB/DC combination or cash balance plan. According to [the Bureau of Labor Statistics], as of 2014, 86 percent of state workers and 82 percent of local workers had access to a DB plan, while only 43 percent and 30 percent had access to a DC plan, respectively.

According to NASRA [the National Association of State Retirement Administrators], about 75 percent of state and local workers participate in Social Security through their current jobs. About 40 percent of all public school teachers don’t participate in Social Security in their current roles, and about two-thirds of public safety workers don’t participate in Social Security. There are some states, such as Alaska, Colorado, Louisiana, Maine, Massachusetts, Nevada, and Ohio, where essentially all public employees don’t participate in Social Security.

A couple of states—Alaska and Michigan—offer only a DC plan for state workers. Eight states offer a hybrid plan that is either a combination of DB and DC or a cash balance plan. Nine states offer workers a choice between plan types. There has been growing interest in the hybrid plan design.

He goes on to explain the increasing importance of DC plans:

Changes to DB as well as to health care plans increase the importance of supplemental DC savings in the state and local sector. While we refer to DC as “supplemental savings,” these plans are a growing part of building retirement security. For workers to retire on time and support a reasonable standard of living, they will likely need a DC plan, other personal savings, and Social Security, if they participate. While there is no set retirement income objective, many folks are thinking 85 percent of final pay is an appropriate goal.

WHAT IS A REASONABLE PAY REPLACEMENT TARGET?

Several factors influence the pay replacement target, including the public pension (e.g., Social Security) amount, availability of a DB plan, worker income level, tenure, and assumed retirement age. We are often asked, “How do you have a target income replacement percentage for a population that is highly diverse?” While imperfect, focusing the plan to the median worker may be most appropriate. Workers with lower incomes may have more of their pay replaced by Social Security, while higher-income individuals may have other compensation programs or personal savings to help meet retirement income needs. It is also important to note that the overall income replacement level will vary by individual and income band. Those with the highest income may need a far lower income replacement level, while those with the lowest income may need more.

While beyond the scope of this book, retiree medical cost coverage is also a significant factor to consider. Only a small percentage of U.S. private employers cover retiree medical costs, and this number continues to decline according to unpublished Employee Benefit Research Institute estimates from the Medical Expenditure Panel Survey: “Very few private-sector employers currently offer retiree health benefits, and the number offering them has been declining. In 2014, 16.6 percent of workers were employed at establishments that offered health coverage to early retirees, down from 28.9 percent in 1997.”

By contrast, many public employers do cover retiree medical, although according to Franzel this is beginning to change:

A few public employers have dropped retiree medical altogether or changed to a DC model for retiree health coverage. Unlike DB plans that may be protected by state constitutions or statute, health care benefits typically do not have the same protections. As a result, health care benefits are more likely to be cut. That’s another reason why DC plan availability and contributions will likely grow—retirees may look to their DC plan to help fund more of their retiree medical expenses.

Workers outside the United States are far more likely to have retiree medical coverage via the government. Given that the majority of U.S. workers lack both a retiree medical and a DB plan, we asked consultants in our 2016 Defined Contribution Consulting Support and Trends Survey: What total income replacement percentage should the median worker seek? At the median, consultants suggested 80 percent as the total replacement target, yet some suggested as little as 50 percent and others as much as 120 percent. That’s quite a range. Let’s look at income sources that can help fill this need.

Public pension income such as Social Security is the first building block in real retirement income replacement. Today, U.S. Social Security provides 35 percent of retirees’ income. In the OECD’s Pensions at a Glance, 2015, they provide expected public pension income replacement by country for various worker income levels. As shown in Figure 2.2, the public pension expected income replacement level for the median worker ranges from 4 percent in Mexico to 70 percent in Italy. This means DB, DC, or personal savings must fill the gap.

FIGURE 2.2 Expected Retirement Income from Public Pension

Market Funded Pension System Gross Pension Replacement Rates
Country Structure Public Mandatory Private Voluntary DC Total
Australia Mandatory DC 14% 31% 0% 45%
Canada Voluntary DC/DB 37% 0% 29% 66%
Denmark Mandatory DC 22% 46% 0% 68%
France Voluntary DC/DB 55% 0% 0% 55%
Germany Voluntary DC/DB 38% 0% 13% 51%
Italy Auto-enrollment DC 70% 0% 0% 70%
Japan Voluntary DC/DB 35% 0% 0% 35%
Mexico Mandatory DC 4% 22% 0% 26%
United Kingdom Auto-enrollment DC 22% 0% 30% 52%
United States Voluntary DC/DB 35% 0% 33% 68%

Source: OECD, Pensions at a Glance 2015.

The OECD also projects what percentage of pay replacement may come from occupational pension programs, such as DC plans. Figure 2.2 shows that countries may have mandatory or voluntary DB or DC systems. For instance, Australia requires participation in a DC plan (i.e., superannuation), whereas the UK requires auto-enrollment in a DC plan. The difference in the latter case is that participants may decide to opt out of the DC plan. Based on this OECD data, U.S. plan sponsors may consider an income replacement target from DC of 40 percent or more. Consultants place this target for U.S. workers higher at 60 percent, yet some suggested as low as 20 percent.

Plan sponsors also may vary the replacement income target based on the demographic characteristics of their organization’s population. For instance, airline pilots may have shorter careers but higher pay than employees in many organizations. David Fisser, consultant and former chairman of the Southwest Airline Pilots’ Association 401(k) Committee in a July 2009 DC Dialogue said this about pilots: “They have higher incomes. They also have a mandatory retirement age of no later than 65 years, which works well because our average new-hire age is about 35 now. If participants can work until they’re 65, they have around 30 years to accumulate enough funds to provide a decent retirement.”

By comparison, university professors may have longer careers but possibly lower pay levels. Yet, the pilot may not have a DB plan while the professor may. As a final example, retail organizations may have both short tenure and lower pay levels—and no DB plan.

As we consider an appropriate income replacement target, plan sponsors should also consider what would be considered failure. For instance, if the target is to replace 60 percent of final pay, would replacing only 30 percent be considered failure? As we consider DC design throughout this book, we use as an example income replacement target of 50 percent and consider achieving less than 30 percent as failure. Plan sponsors may set higher or lower targets and apply the same evaluation methods to determine likely success.

CALCULATING THE INCOME REPLACEMENT RATES

How do plan fiduciaries calculate the replacement rate historically as well as prospectively? While few participants actually buy an annuity, a common approach to calculating income replacement ratios is to consider the cost of buying a lifetime income stream via an immediate annuity.

For example, consider a 65-year-old male who has accumulated $400,000 in DC assets. This is a substantial sum, but is it enough to deliver an adequate income in retirement? One way to answer the question is to look at the income that could be delivered by purchasing an annuity. Based on average annuity quotes between March 2013 and December 2015 provided by Hueler Companies (Figure 2.3), he may have received an annuity payout equal to a 6.47 to 7.17 percent return, providing an annual income of $25,899 to $28,665 a year. If we assume his final pay was $75,000, his income replacement rate would be 35 to 38 percent, depending on the year in which he retired.

Graph: Payout rate with 0 to 8% versus x-axis with Mar-13 to Sep-15 has solid, dashed horizontal curves from 4-5%, 6-7% of x-axis as real, and nominal payout rate with average being 72%.

FIGURE 2.3 Immediate Annuity Income: Real Payout Rate versus Nominal Payout Rate

Sources: Hueler Companies, Inc., and PIMCO, as of December 31, 2015. Hueler data is © 2015 Hueler Companies, Inc. All Rights Reserved.

But what if he wanted to purchase a real annuity in which the payout is adjusted annually consistent with the Consumer Price Index (CPI)? Over the period from March 1, 2013 to December 31, 2015, real annuity quotes averaged 72 percent of the nominal payout (see Figure 2.3). This means the dollars paid would have delivered $17,885 to $20,501 a year, or 4.47 to 5.13 percent annually, in real terms. His real income replacement rate would then equal 24 to 27 percent of his final salary.

To consider a longer time frame historically or to model prospectively, we need a proxy for annuity pricing. At PIMCO, we developed a methodology for this proxy that we refer to as the PIMCO Retirement Income Cost Estimate (or PRICE). PRICE is calculated as the discounted present value of a 20-year annual income stream using the historical zero-coupon U.S. TIPS yield curve. After considering many alternatives, we have found that a 20-year ladder of zero-coupon TIPS provides the best available proxy for annuity rates. If we compare real annuity rates to the TIPS ladder (which is a portfolio of TIPS with different maturities), we see a 91 percent correlation and a small difference in the payout rate (Figure 2.4).

Graph: immediate annuity payout rate 4-5.2%,TIPS ladder retirement income payout rate 4.4-5.6% versus mar13-sep15 has curves; Table has values for real payout rate, et cetera.

FIGURE 2.4 Correlation of Real Annuity Rates with TIPS Ladder Retirement Income

Sources: Hueler Companies, Inc., and PIMCO, as of December 31, 2015. Hueler data is © 2015 Hueler Companies, Inc. All Rights Reserved.

In our view, a 20-year inflation-adjusted income stream from TIPS makes sense as a proxy for inflation-adjusted income, particularly as we consider life expectancy for a 65-year-old male today is 17.4 years, and for a female 20.4 (according to the 2010 life expectancy tables from the Society of Actuaries). Thus if a retiree bought a 20-year ladder of TIPS bonds, as shown in Figure 2.5, he or she would have real purchasing power over those 20 years. We considered adjusting the TIPS ladder for a mortality credit or otherwise, but found a simple ladder to have the best fit to the actual real annuity pricing.

Graph: annual distribution with 0-25,000 $ versus years in retirement with 1-18 has bars for real income requirement and inflation adjustment from 0 and 20,000 of y-axis respectively.

FIGURE 2.5 Real Income in Retirement from a TIPS Ladder

The chart above depicts the components for the total distribution that an investor receives over time. The cumulative inflation accruals may be adjusted higher or lower as a result of increases (inflation) or decreases (deflation) in the Consumer Price Index (CPI). Due to the laddered structure, the amount of principal in the distributions is higher in later years.

Source: PIMCO. For illustrative purposes only.

HISTORIC COST OF RETIREMENT: PRICE IS A MOVING TARGET

Participants often ask, “How much do I need to retire?” Or “What’s my number?” Unfortunately, this is not an easy question to answer as the cost to retire will depend largely on prevailing interest rates at the time of retirement. Using PRICE helps us look both at the historic cost of retirement as well as the potential future cost.

How much people need for retirement is a moving target. Over the last decade, using PRICE as the definition of the cost of retirement, near-retirees have faced an escalating retirement price tag. This change is shown in Figure 2.6, which graphs the change for a typical participant with final pay of $75,000 and a 30 percent real retirement income replacement goal (i.e., $22,500 CPI-adjusted annual income). As you can see in Figure 2.6, from a 2008 low to a peak in 2013, the near-retiree saw her retirement cost jump by about $150,000—or twice her final pay.

Graph: price 300,000-500,000 $, 10 year real rate 1.50-4% versus x-axis Jan4-Jan15 has fluctuating curves with peaks for price (LHS), 10-year real rate (RHS) with $321,327, et cetera.

FIGURE 2.6 PRICE Is a Moving Target: How Much Does a Participant Need to Replace 30 Percent of Final Pay?

Hypothetical example for illustrative purposes only.

Assumptions: (1) a participant has $75,000 in final salary regardless of the time period; (2) his or her goal is to achieve a 30 percent income replacement rate after retirement.

*The retirement income cost is calculated as the discounted present value of the 20-year annual income stream of $22,500 by using the historical zero-coupon U.S. TIPS yield curve.

Sources: PIMCO, Bloomberg Finance L.P., and Haver Analytics, as of December 31, 2015.

Because the retirement cost is calculated as the discounted present value of future cash flows, it is not surprising to see that it is inversely correlated with interest rates: As interest rates move up, the cost of retirement moves down, and vice versa. This relationship underscores why retirees desperately need rates to rise!

Figure 2.6 only takes us so far, however; as while understanding historical retirement costs is useful, and can provide insight into cost over time, workers and plan fiduciaries likely care more about future costs.

This is where PRICE can help. Figure 2.7 shows the cost of retirement based on a participant’s age, an assumed retirement at age 65, and desired annual income stream, as of December 2015. Looking at Figure 2.7, you can see that at the end of 2015, for a 25-year-old to buy an annual CPI-adjusted income stream of $50,000, the cost would have been just under $530,000. In contrast, for participants age 45 or 65, buying the same real annual income would have cost about $685,000 and just over $900,000, respectively. The discount rate (used to determine the present value) differs depending on the years to retirement. The closer a participant is to retirement, the higher the retirement cost.

FIGURE 2.7 Cost for Future Retirement Income Stream

As of December 31, 2015 Cost for Future Annual Retirement Income Stream
Age PRICE Multiplier $25,000 Annual Income $50,000 Annual Income $100,000 Annual Income
25 10.59 $264,794 $529,589 $1,059,178
35 12.02 $300,471 $600,942 $1,201,884
45 13.70 $342,559 $685,118 $1,370,237
55 15.79 $394,726 $789,453 $1,578,906
65 18.15 $453,836 $907,671 $1,815,343

Hypothetical example for illustrative purposes only. Multiplier is rounded so annual income amounts differ.

Sources: PIMCO, Bloomberg Finance L.P., and Haver Analytics, as of December 31, 2015.

To consider the cost of different annual income streams, readers may use the “PRICE multiplier” shown in Figure 2.7. The PRICE multiplier is the cost of one dollar of annual retirement income, and is calculated as the discounted present value of the 20-year zero-coupon TIPS ladder. It shows, for example, that for a 25-year-old seeking $100,000 of real annual income in retirement (with an assumed retirement at age 65—meaning they are 40 years from retirement, 20 years from retirement, or have reached retirement), the purchase price was simply the annual retirement income times the PRICE multiplier, or approximately $1.06 million. Figure 2.8 shows the PRICE multiplier change over time (January 2004 to December 2015) for participants at 25, 45, and 65 years old.

Graph: price multiplier with 0-25 versus x-axis Jan4-15 has horizontal fluctuating curves for at retirement, 20 and 40 years to retirement; Table: has values for price multiplier, et cetera.

FIGURE 2.8 Price Multiplier for Different Age Groups

Hypothetical example for illustrative purposes only.

Sources: PIMCO, Bloomberg Finance L.P., and Haver Analytics, as of December 31, 2015.

As you can see in Figure 2.8, since inception in February 2004, the annualized change in the PRICE multiplier is highest for the 40 years to retirement vintage at 3.7 percent and lowest for the at-retirement vintage at 0.9 percent. This change is largely a function of the longer duration of retirement liability and the declining interest-rate environment.1 Looking at the volatility of the PRICE metric, from inception to December 2015, the PRICE multiplier had annualized volatility of 7.3 percent at retirement (i.e., age 65). PRICE multiplier volatility is much higher 20 and 40 years prior to retirement at 20.4 percent and 34.7 percent, respectively. What this means for prospective retirees is that a considerable amount of uncertainty remains in retirement planning as the years from retirement increase and uncertainty about the prospective cost of retirement declines as the retirement date approaches.

A FOCUS ON INCOME, NOT COST

Some participants may flip the “What’s my number?” question and ask, “How much income can I buy given my current savings?” Similarly, plan fiduciaries may consider what percentage of income can be replaced based on the median balance in their DC plan.

The PRICE multiplier can help give insight to these questions as well. Figure 2.9 shows how much income (and what replacement rate) various levels of savings could confer as of December 2015. For a 65-year-old with $500,000, the multiplier is 18.15, leading to a projected annual income of about $27,500—or a 37 percent income replacement rate, assuming her final pay is $75,000. If her accumulated balance increases to $700,000, the income replacement rate increases to 51 percent.

FIGURE 2.9 Estimated Income Replacement from Various Levels of Savings

Accumulated Balance at Age 65 PRICE Multiplier Annual Income Income Replacement Rate
$500,000 18.15 $27,543 37%
$600,000 18.15 $33,052 44%
$700,000 18.15 $38,560 51%
$800,000 18.15 $44,069 59%
$900,000 18.15 $49,577 66%
$1,000,000 18.15 $55,086 73%

Assumes a participant has $75,000 in final salary regardless of the time period. Hypothetical example for illustrative purposes only.

Sources: PIMCO, Bloomberg Finance L.P., and Haver Analytics, as of December 31, 2015.

PRICE-AWARE: APPLYING PRICE TO CONSIDER DC ASSETS AND TARGET-DATE STRATEGIES

Now that we have identified PRICE as a proxy for the historic and future cost of retirement (i.e., the real liability), we can evaluate assets and asset allocation structures relative to PRICE. To reduce risk relative to the retirement-income replacement objective, DC plans—like their DB cousins—should seek assets that are liability-aware. PRICE tells us the cost of the DC liability. If we look at asset classes, we can determine which ones correlate best to the liability. Figure 2.10 shows a low correlation of 0.11 between the monthly returns of the S&P 500 Index and the retirement liability from January 2004 through December 2015. In contrast, Figure 2.11 shows the correlation between the retirement liability and long-duration TIPS is very “tight” (0.95).

Graph: Return with -20 to 15 % versus x-axis with Feb4 to Oct15 has fluctuating curves with peaks from 0-5% of y-axis for S and P 500, retirement liability with correction as 0.11.

FIGURE 2.10 S&P Index versus Retirement Liability

Retirement liability represented by a 20-year zero-coupon TIPS ladder.

Sources: Bloomberg Finance L.P., Haver Analytics, and PIMCO, as of December 31, 2015.

Graph: return with -20 to 15% versus x-axis with Feb4 to Oct 15 has fluctuating curves with peaks from 0-5% of y-axis for long TIPS and retirement liability with correction as 0.95.

FIGURE 2.11 Long-Duration TIPS versus Retirement Liability

Retirement liability represented by a 20-year zero-coupon TIPS ladder. Long TIPS represented by Barclays U.S. TIPS Index > 10 Years.

Sources: Bloomberg Finance L.P., Haver Analytics, and PIMCO, as of December 31, 2015.

In Figure 2.12, we show the absolute risk (volatility) side-by-side with the risk relative to retirement income generated by the TIPS ladder. This illustrates the reduced retirement income risk of long-duration TIPS compared to stocks, nominal bonds, and even cash.

Bar graph: volatility with 0-16% versus x-axis has paired bars on x-axis for BofA ML 3-month U.S Treasury bill index, Barclays capital U.S aggregate index, with 0.7%, 7.3%, et cetera.

FIGURE 2.12 Considering Absolute Risk of Selected Assets versus Risk Relative to Retirement

Absolute risk is measured by the standard deviation of monthly returns of the asset class from February 2004 to December 2015; relative risk to retirement liability is measured by the standard deviation of surplus returns over retirement liabilities (represented by a 20-year zero-coupon TIPS ladder) from February 2004 to December 2015.

Sources: Bloomberg Finance L.P., Haver Analytics, and PIMCO, as of December 31, 2015.

Selecting assets that correlate with PRICE better aligns the investments to the DC plan objective. Unlike a DB plan, DC plans are unlikely to fully match the assets to the liabilities. In the February 2011 DC Dialogue, Professor Zvi Bodie shared with us a basic formula which shows that if participants want to replace 50 percent of their income in retirement and desire to retire at age 65, they need to begin saving 25 percent of their pay at age 25, assume a zero real interest rate on their savings, and expect to live to age 85. In reality, on average the DC participants’ saving rates are far below 25 percent, according to 2016 EBRI Retirement Confidence Survey. DC plan fiduciaries would not buy only TIPS—rather, they will likely include other assets that may offer additional return and accept the risk (i.e., tracking error) relative to PRICE. Target-date funds offer the most prevalent example of a DC asset allocation structure. In Chapters 3 and 4 we will take a close look at investment structures and target-date strategies. For now, let’s take a quick look at how an asset allocation glide path may be evaluated relative to PRICE.

EVALUATING GLIDE PATHS

By aligning the target-date design to the retirement liability PRICE, DC participants have a higher likelihood of meeting their income goals. Let’s consider two glide paths: the Market Average Glide Path (which is constructed by NextCapital and is an average of the 40 largest target-date strategies in the market) and an Objective-Aligned Glide Path, as shown in Figure 2.13.

Graph: allocation with 0-100% versus years of retirement with 40-0 has horizontal curves with different shades for glide path asset allocation. Graph: allocation with 0-100% versus years of retirement with 40-0 has curves with different shades for fixed income and cash, inflation linked, stock, et cetera.

FIGURE 2.13 Glide Path Asset Allocation

Market Average data is as of September 30, 2015. Objective-aligned data is as of December 31, 2015.

Sources: PIMCO and NextCapital.

The Objective-Aligned Glide Path has a greater proportion of assets allocated to TIPS, including an allocation of between 1 percent and 10.0 percent to long-duration TIPS. Back-tested results from January 2004 to December 2015 (see Figure 2.14) show that the Objective-Aligned Glide Path at-retirement vintage has a tighter correlation to the retirement liability at .51 than the Market Average Glide Path at-retirement vintage at .38.

Graph: correlation to retirement liability with -1 to 1 versus x-axis with Jan 7-15 has horizontal fluctuating curves for objective aligned and market average glide path, et cetera.

FIGURE 2.14 Correlation between Glide Paths and Retirement Liability

Retirement liability represented by a 20-year zero-coupon TIPS ladder.

Sources: Bloomberg Finance L.P., Haver Analytics, NextCapital, and PIMCO. Market Glide data is as of September 30, 2015; Objective-Aligned Glide Path data is as of December 31, 2015.

Plan fiduciaries may worry that a tighter correlation relative to PRICE may result in lower investment returns. This is not necessarily the case. According to back-tested results from February 2004 to December 2015, the Objective-Aligned Glide Path yields greater inflation-adjusted capital appreciation. For instance, assume participants started with $350,000 and invested in the Market Average Glide Path and the Objective-Aligned Glide Path at age 55 (from January 2004 through December 2015). Figure 2.15 shows that the accumulated real balance is $42,751, or 6.1 percent higher, for the Objective-Aligned Glide Path compared to the Market Average Glide Path.

Graph (top): y-axis 300,000-900,000 $ versus x-axis Jan 4-15 has fluctuating curves, et cetera; Table has values for rows with A, B, A/B, columns for objective aligned glide path, et cetera.

FIGURE 2.15 Accumulated Real Balance

Hypothetical example for illustrative purposes only. Salary at age 55: $67,000; real annual wage growth: 1 percent; savings rate: 12 percent; employer match: 3.5 percent.

Sources: PIMCO, NextCapital, Bloomberg Finance L.P., and Haver Analytics, as of December 31, 2015.

As fiduciaries evaluate DC plan investments, they will need to bring together both risk and return considerations. As the most prevalent DC investment default, target-date funds rise to the top in requiring scrutiny and benchmarking. We suggest plan sponsors evaluate target-date strategies relative to their investment objective, which consultants define as “to maximize asset returns while minimizing volatility relative to the retirement liability.”

Evaluating asset structures relative to this objective can be done using an information ratio, that is, a risk-adjusted return measure calculated using active return divided by tracking error, where active return is the difference between the return of the security and the return of a selected benchmark index, and tracking error is the standard deviation of the active return. Figure 2.16 shows how target-date glide paths can be evaluated relative to PRICE within an information ratio context. Again, comparing the two glide paths—the Market Average Glide Path and the Objective-Aligned Glide Path—we see in Figure 2.16 that across all “vintages” (years in which income will start to be required from the portfolio—such as the at-retirement vintage, or the 20-years-to-retirement vintage), the Objective-Aligned Glide Path has a higher excess return, lower tracking error, and higher information ratio. The at-retirement vintage shows a 0.72 information ratio for the Objective-Aligned path compared with 0.60 for the Market Average path.

Graph: excess return 0.00-6.00% versus tracking error 0-40.0% has declining curves for objective aligned and market average glide path; Table has values for information ratio, et cetera.

FIGURE 2.16 Excess Return versus Tracking Error to Price (February 2004– December 2015)

Hypothetical example for illustrative purposes only.

Sources: PIMCO, Bloomberg Finance L.P., and Haver Analytics, as of December 31, 2015.

This assessment tells fiduciaries whether participants have been compensated for taking on the added risk, or tracking error, as measured by the information ratio (which for our purposes is the ratio of excess return to tracking error, both relative to PRICE). The downward slope that you can see in Figure 2.16 is explained primarily by the lower excess return and higher tracking error over PRICE of the farther-dated vintages (i.e., 20 or more years to retirement). Keep in mind that the slope will change over time based on asset returns and the interest-rate environment. Looking at Figure 2.13, the overall story is that the Objective-Aligned Glide Path has been a better choice for participants than the Market Average Glide Path, when compared on the metrics of excess return, tracking error, and information ratio.

From Figure 2.13, we know that the Objective-Aligned Glide Path has a higher allocation to liability aware assets like long TIPS than the Market Average Glide Path. The level of real rates at retirement has a profound effect on income replacement. An allocation to long duration assets late in the glide path can help mitigate the likelihood of achieving a very low income replacement ratio during a low rate environment. Figure 2.17 shows that compared to the Market Average Glide Path, the Objective-Aligned Glide Path has lower probability of having an income replacement ratio less than 30 percent under different real rate environments and the difference is larger when real interest rate is lower.

Graph: Probability of failure with 0-70% versus real rate at retirement with -1.0-3% has paired bars in decreasing order for market average and objective aligned glide path.

FIGURE 2.17 Probability of Failure (Income Replacement Ratio <30 Percent)

Hypothetical example for illustrative purposes only.

Sources: PIMCO, NextCapital, as of December 31, 2015.

TRACKING DC ACCOUNT BALANCE GROWTH RELATIVE TO PRICE

Taking our analysis further, plan sponsors may also want to evaluate how the default investment strategy performs relative to PRICE by taking into account the impact of savings.

Here’s how it works. Assume a participant at age 55 has accumulated $350,000 in retirement savings. Her personal savings rate and the employer match from age 55 through 65 are 12 percent and 3.5 percent per year, respectively. Her salary at age 55 is $67,000 and the real annual wage growth rate is 1 percent per year. Given this information, we can calculate and compare the DC account balance that would have accumulated by investing in the Objective-Aligned Glide Path and the Market Average Glide Path, both relative to PRICE.

Figure 2.15 tells us that the average deviation from PRICE in the accumulated balance has been positive for the Objective-Aligned Glide Path and negative for the Market Average Glide Path between January 2004 and December 2015. Although both glide paths outpaced PRICE—they ended ahead of the retirement cost—from the perspective of relative risk, the Objective-Aligned Glide Path had significantly lower tracking error in its accumulated balance than the Market Average Glide Path. Thus, the Objective-Aligned Glide Path had a positive ratio of average deviation to tracking error in accumulated balance at 0.56, whereas the corresponding ratio for the Market Average Glide Path was –0.11. Again, the overall story is that the Objective-Aligned path presented a more successful outcome for participants over this timeframe.

SUMMARY: THE IMPORTANCE OF KNOWING YOUR PRICE

As with reaching any goal, a concrete objective—knowing one’s number—can be helpful. Although it is a theoretical construct, not a prediction or projection of investment return, PRICE can help in this regard. It can translate accumulated account balances into future retirement income potential. And it can help plan sponsors benchmark target-date strategies.

Plan fiduciaries, especially those in the United States, will want to keep a close eye on how their target-date funds have performed, and how likely they are to deliver the retirement income plan participants need. PRICE is a simple yet compelling methodology to address this most fundamental of retirement planning questions. We believe that by helping to keep track of and projecting one’s progress in real terms, the PRICE metric can be an invaluable aid for individuals and plan sponsors alike.

IN CLOSING

In the first and second parts of this chapter, we’ve taken on several of the biggest questions about DC plans; questions that ask and seek to answer how plans might be best positioned to deliver the retirement income workers need. They are:

  • What might retirement cost, and what kinds of plans provide the best shot at meeting the PRICE of retirement?
  • How can plan fiduciaries evaluate default investment options in DC plans—particularly target-date strategies—and determine which default options are most likely to allow retirees to retire, both on time and with sufficient lifetime income?

In the last section, we also looked beyond the quantitative aspects of plan development and performance to consider the qualitative aspects, or human factors that influence and shape plan design—and we’ve drawn on the expertise of knowledgeable observers to suggest ways in which plans can take those factors into account to deliver a higher chance of meeting plan, sponsor, and retiree objectives.

QUESTIONS FOR PLAN FIDUCIARIES

  1. What is the investment objective for your DC plan?
  2. How is the plan design aligned to this objective?
  3. What does it cost to retire?
  4. Are your DC investments aligned to the retirement cost?
  5. What assets may better align to the cost?
  6. How much tracking error are you willing to take versus the cost?
  7. Are you benchmarking to the cost?

NOTE

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset