Proposition Three

By Craig L. Israelsen, Ph.D.
Originally published in Financial Planning Magazine, November 2007

f you've ever tried to evaluate and compare target-date funds, you've no doubt realized this is no simple task. There are numerous interpretations of what constitutes a prudent and appropriate asset allocation over the lifetime of the investor, and each fund seems to approach this issue differently. Consequently, there is a need for independent target-date indexes that can serve as performance benchmarks for target-date funds. To date, the only suite of target-date indexes available is the Dow Jones Portfolio Target series. This article seeks to address this gap by proposing a methodology for the construction of a set of target date indexes which incorporates several prudent (i.e., "best practice") elements that take into account the need for adequate capital growth and the need for safety of principal.

But before we delve into the particulars of these proposed indexes, let's review the general strategy behind target-date funds. The primary purpose of a target-date fund is to replicate "best practice" transitions in a portfolio's asset allocation strategy over the lifetime of an investor. This transition process through time is often called the glidepath. The primary purpose of a glidepath is to adjust the risk characteristics of a portfolio over time. When the portfolio is 30 to 40 years away from the target date it is weighted to provide a higher return through higher risk. As the fund approaches the target date the glidepath calls for allocations that emphasize lower risk, and hence offer a lower potential return.

It is not possible to determine, in advance, which glidepath and/or asset allocation model will be optimal among myriad target-date funds presently available. That will only be known far down the road as we look back at historical performance. Nevertheless, this article introduces three target-date indexes: Conservative Pure Target Index, Moderate Pure Target Index and Aggressive Pure Target Index. Each index has six target dates: Current, 2010, 2020, 2030, 2040 and 2050. These target-date indexes, developed in collaboration with Joe Nagengast and Ron Surz, represent a departure from the typical glidepath guiding most target-date funds available today. We present these indexes as a blueprint from which a new breed of target-date indexes and/or target-date funds might be built.

One of the primary challenges in developing target-date indexes for use as benchmarks to evaluate target-date funds is the wide variation that exists among these funds, even within the same target cohort (i.e., 2010 funds, 2020 funds, etc.). This is precisely why we developed three target-date index categories: Conservative, Moderate and Aggressive. Having three distinct target-date index categories-that share the same pedigree in terms of theoretical design and core portfolio holdings, but with different glidepaths in the 20 years prior to the target date-will allow for more meaningful benchmarking of individual target-date funds. Rather than simply comparing the performance of the XYZ 2020 Fund against the average performance of all 2020 funds, the benchmarks make it possible to determine if the XYZ 2020 Fund is a conservative, moderate or aggressive 2020 fund. Just as other mutual funds have their best-fit index, target-date funds will have their best-fit Pure Target Index within each target-date cohort.

As shown in the three figures ("Conservative Pure Target Index," "Moderate Pure Target Index" and "Aggressive Pure Target Index"), the glidepaths in these target-date indexes span a 40-year period, simulating only the accumulation period for an investor. The accumulation period ends at the target date, which typically coincides with the retirement year of the investor. A 40-year accumulation period may seem overly optimistic given that far too many people don't start investing in their mid-twenties. However, auto enrollment of employees into retirement plans should begin to remedy this problem. Add to that the likelihood that today's young workers will not likely retire until their late sixties or early seventies, and a 40-year accumulation period begins to appear quite likely.

Risky Business

How are the Pure Target Indexes different from the typical target-date fund design? One aspect is that we use only two broad investment classes: a risky asset group and a reserve asset group. The reserve asset group is comprised of assets that protect principal.

The risky asset group is comprised of six separate subassets: two broad equity assets that have US and non-US exposure, two broad US and non-US bond assets, and two diversifying assets that focus on asset classes that tend to have low-correlation to the other assets in the portfolio. Collectively, these six risky assets represent a capitalization-weighted, global basket of investable assets.

It's instructive to note that we include two bond assets and two equity diversifier assets as components of the risky asset group. Including bonds as a "risky" asset is a departure from the norm, but is in keeping with the central tenets of Markowitz's efficient frontier (which includes all risky assets). However, missing from the Markowitz efficient frontier is cash (and other reserve assets). Enter the Capital Market Line, a logical extension of the efficient frontier into non-risky assets. By including reserve assets in the Conservative and Moderate portfolios, it is possible to create a portfolio mix that has risk/return characteristics that are superior to a portfolio that includes only those risky assets that reside on or near the efficient frontier. (The inclusion of reserve assets only pertains to the Conservative and Moderate Indexes).

Another difference between our indexes and the typical target-date fund in that in the Conservative and Moderate Pure Target Indexes we implement a transfer-of-assets protocol (through annual rebalancing) which creates a "lockbox". The lockbox insulates money from loss caused by equity market gyrations by migrating portions of the portfolio into the reserve asset group at an increasing rate as the investor approaches the target date. In fact, at the target date (the anticipated retirement year of the investor), the allocation of the Conservative Index is 100% in reserve assets and the allocation of the Moderate Index is 50% in reserve assets. While not guaranteed like an annuity, this lockbox concept (advocated by Bill Sharpe) incorporates elements of liability driven investing (LDI) into the construction of a target-date glidepath. Liability driven investing is not the mantra of most target date funds in the market today. At retirement, the investor is then able to determine how best to position his or her portfolio. We anticipate that many individuals will choose to purchase an annuity product after the target date is reached. We are suggesting that LDI represents a "best practice" that should be followed when building a conservative or moderate target-date fund or target-date index.

It is our opinion that target-date funds comprise two very separate functions: an accumulation phase and a decumulation (or distribution) phase. Both phases present different goals and challenges unique to each. We have chosen to solely focus on the accumulation phase in our Pure Target Indexes. We suspect that in the future, target-date funds may be segmented into two separate products: an accumulation vehicle and a distribution vehicle. These Pure Target Indexes represent accumulation indexes, rather than indexes that attempt to cover both the accumulation and distribution periods.

Compare and Contrast

How do the Pure Target Indexes compare against the Dow Jones Portfolio Target Indexes and the average performance of all target-date funds? The table "On Target," lists the annual returns from 1998 to 2006, as well as the nine-year annualized return, the nine-year standard deviation of annual returns and the return/risk score. The return/risk score is a simplified Sharpe ratio where the higher the score the better. Understandably, the number of actual target-date funds comprising the average performance number was smaller in 1998 than in 2006. In fact, nearly 60% of the target funds currently available came into existence after January 1, 2005.

The annual returns used are based on the actual historical performance of the specific underlying assets in each Pure Target Index, as determined by the asset allocations specified in the glidepath (dependent upon years to target date). In the table, the term "Current" indicates that the target date has been reached. (Some current funds interchangeably use the titles income or today.) The raw data used in this study were obtained from Morningstar.

Among "Current" funds and indexes, the Pure Target Conservative Index (lavender) had considerably better returns during the difficult years of 2000 to 2002. That is a direct result of the previously mentioned lockbox concept. The nine-year annualized return of the Pure Target Conservative Index (5.7%) was lower than the 6.5% of the Dow Jones Portfolio Target Today Index (light blue), and slightly lower than the 5.9% average annualized performance of all target-date funds (yellow). But note that the return/risk score was highest for the Pure Target Conservative Index. As would be anticipated because of the larger equity exposure, the nine-year annualized return for the Pure Target Moderate Index (7.6%) and Pure Target Aggressive Index (9.0%) were higher than both the DJ Today Index and the average performance of all current target funds. The price of higher performance is higher volatility, as noted by the considerably higher standard deviation of return in the Moderate and Aggressive Pure Target Indexes.

Among 2010 funds and indexes, the Pure Target Conservative Index generated the highest return/risk score (1.04). This is directly attributable to its loss resistance in 2000 and 2002. The nine-year performance and risk of the Dow Jones Portfolio Target 2010 Index fell in between the performance of the Pure Target Conservative and the Pure Target Moderate Indexes. The average performance of all 2010 target funds fell short of all the indexes, but the volatility of return (6.7%) was in line with the Pure Target Conservative Index. (You will note that the performance of the Pure Target Aggressive Index is identical in each target-date cohort because it has constant allocations during the accumulation period.)

All three Pure Target Indexes shine in the 2020 cohort, generating higher annualized returns and return/risk scores than the DJ 2020 Index and the average nine-year return of all 2020 funds. As before, this is largely a function of their performance in 2000 and 2002. The nine-year volatility of return in the Pure Target Moderate 2020 and Pure Target Aggressive Index 2020 was slightly higher than the DJ 2020 Index and the average 2020 target fund. In 2001, the Pure Target Indexes underperformed to the DJ 2020 Index.

The year-to-year performance of Pure Target 2030, 2040 and 2050 Indexes were identical within each risk category (Conservative, Moderate, and Aggressive). The 2030, 2040, and 2040 Pure Target Conservative average annual return was 8.6% and a 12.5% standard deviation. The Pure Target Moderate Indexes for 2030, 2040, and 2050 had an average annual return of 8.8% with a 13.7% standard deviation, while the Pure Target 2030, 2040, and 2050 Aggressive Indexes had an average annual return of 9.0% and a 14.9% standard deviation of return. The performance of the three Pure Target Indexes in these distant target dates (2030, 2040, 2050) is not a result of the glidepath, because the glidepath remains flat during the early years. Rather, the difference in performance in the early years (more than 20 years prior to target date) is a direct result of the slightly different asset allocation model being utilized within each portfolio. Recall that each index series (Conservative, Moderate, Aggressive) utilizes the same assets. The only exception is that the Aggressive Index does not utilize reserve assets.

The DJ 2030 and 2040 Indexes were averaged together, resulting in a nine-year annual return of 8.6%, but with a standard deviation of 17.1%. The average performance of all target funds in the cohorts of 2030 through 2050 was 6.6% with a standard deviation of 16.1%. Thus, the return/risk scores for the Pure Target Indexes were considerably higher in these distant target date cohorts. The pattern of superior loss resistance for the Pure Target Indexes during 2000 and 2002 was again manifested.

Design Virtues

A distinct advantage of the Pure Target Indexes is their enhanced resistance to loss. This is achieved by the glidepath and/or by the asset allocation model being utilized in the three indexes. It's important to remember that our target-date indexes (and target funds in general) have two vitally important elements: a core asset allocation model and a glidepath design. In our Pure Target Indexes, performance during the first 20 years is dictated entirely by the asset allocation model, whereas the performance of the indexes during the 20 years prior to the target date is heavily influenced by the design of the glidepath (in the Conservative and Moderate Indexes).

This study illustrates the virtues of a target-date fund design that incorporates two essential aspects: prudently grow money and progressively protect it as the target date is approached. The Dow Jones Portfolio Target Indexes demonstrate a commitment to both of these mandates. The Pure Target Index asset allocation model-in conjunction with its lockbox-oriented glidepath-represents an additional blueprint for target-date design that also supports both mandates. It is our contention that avoiding losses is the number one priority as a target portfolio approaches its target date.

In addition to being a blue-print for the design of target date indexes and/or target date funds, the Pure Target Indexes will facilitate more accurate benchmarking of target date funds. Currently the tools available to categorize target date funds are nebulous at best. With only one target date index currently in the market (Dow Jones), it is difficult to clearly determine if a particular target date fund (within a particular target cohort, say 2020) is conservative, moderate, or aggressive. The Pure Target Indexes solve that.

The number of target-date funds and the assets flowing into them will dramatically increase in the coming years. As a result, additional target-date indexes will necessarily emerge and evolve. As the first suite of target-date indexes to market, Dow Jones is to be commended for adopting a target-date methodology (developed by Global Index Advisors) that emphasizes risk control and capital preservation, rather than succumbing to the trend that is increasingly evident in target-date space today-equity-heavy portfolios focused on outperformance. The Pure Target Indexes have the same mantra: Prudently grow money and then aggressively protect it as the target date approaches.

Conservative Pure Target Index

As the target date comes closer, the index shifts from a mix of six "risky" assets to 100% reserve assets.

Moderate Pure Target Index

As the target date comes closer, the index shifts from a mix of six "risky" assets to a mix of "risky" and reserve assets.

Aggressive Pure Target Index

Allocations in the index remain constant, and do not become more conservative near the target date.

On Target

Comparison of the Pure Target Indexes' performance to that of the Dow Jones Portfolio Target Indexes as well as the average performance of all target-date funds for different target dates.

1998 - 2006	Annual % Return									Nine-Year Average Annual Return (%)	Nine-Year Std. Dev. of Annual Return (%)	Return/ Risk Score   (higher is better)
	1998	1999	2000	2001	2002	2003	2004	2005	2006
Current
Pure Target Conservative: Current	4.5	3.4	10.4	6.3	10.7	5.2	5.5	3.0	2.3	5.7	3.0	1.87
Pure Target Moderate: Current	8.5	11.4	3.5	-1.2	-0.5	18.9	11.7	6.8	11.3	7.6	6.4	1.19
Pure Target Aggressive: Current	12.6	19.3	-3.4	-8.7	-11.7	32.5	18.0	10.6	20.2	9.0	14.9	0.61
Dow Jones Portfolio Target Today	11.3	2.3	5.2	3.7	8.7	12.5	6.2	2.6	6.5	6.5	3.6	1.79
Avg. for All Current Target-Date Funds	12.4	6.1	5.4	1.4	-0.8	11.1	6.4	3.8	8.3	5.9	4.3	1.39
2010
Pure Target Conservative: 2010	10.8	12.6	1.6	-2.2	-0.3	18.0	10.6	5.3	7.6	6.9	6.6	1.04
Pure Target Moderate: 2010	11.7	16.0	-0.9	-5.5	-6.0	25.2	14.3	7.9	13.9	8.0	10.6	0.76
Pure Target Aggressive: 2010	12.6	19.3	-3.4	-8.7	-11.7	32.5	18.0	10.6	20.2	9.0	14.9	0.61
Dow Jones Portfolio Target 2010	10.7	15.1	-0.4	-0.7	-2.1	21.4	10.8	3.8	7.9	7.1	8.0	0.89
Avg. for All 2010 Target-Date Funds	14.4	13.4	2.5	-2.5	-2.6	14.1	7.5	4.1	8.8	6.4	6.7	0.96
2020
Pure Target Conservative: 2020	12.6	15.9	-2.3	-6.7	-7.1	27.4	15.2	7.9	15.5	8.1	11.8	0.69
Pure Target Moderate: 2020	12.6	17.6	-2.9	-7.7	-9.4