BIPY for Growth and MIPY for Value

he November 2004 Financial Planning Magazine article "The Middle Child" defined and analyzed five portfolio management strategies utilizing Wilshire small cap, mid cap, and large cap growth and value equity indexes over a 25 year period. The best outcome resulted from annually selecting the prior year's middle performing value index, or what we refer to as the "Middle Index Prior Year" (or MIPY) strategy.

In the May, 2005 issue of Financial Planning we further explored the strategy which helped deliver this performance among value indexes (see "Gaming The System"). This current article explores the same five strategies, but among growth indexes only and then concludes with a summary of the growth and value strategies.

When using only growth indexes, the best outcome in the November article came from annually selecting the prior year's best performing index. We call this growth index strategy the "Best Index Prior Year" (or BIPY) strategy. However, only one of the five growth strategies (in the November article) was better than the worst of the five value strategies; the other four value strategies clearly beat all five growth strategies.

This article will demonstrate that using the annual Wilshire growth indexes as the reallocation "signal" (i.e. using Wilshire data to determine which index - large, mid, or small - was the best index in the prior year) and subsequently using Russell ETFs as the actual investment vehicles is a feasible way to implement the BIPY strategy. Over the past 18 years, the average annualized return from this strategy was 11.98% with Russell growth (large cap, mid cap, and small cap) indexes and 12.21% with Dow Jones growth indexes. The returns discussed in this article are pre-tax and exclude annual fees and transaction costs. However, with ETFs, these expenses are very low. Annual reallocation should of course take advantage of capital gains tax rates, or preferably take place within a tax-sheltered account (e.g. Roth IRA, 401(k), etc.). As always, past performance is no guarantee of future results.

The BIPY strategy requires an annual reallocation decision based on selecting the prior year's best performing index. For this type of portfolio reallocation strategy, an index family needs to have a full complement of style and market-cap specific indexes. Perhaps the most ubiquitous cap/style version of the U.S. equity market is the nine-box grid used by Morningstar. The six most distinct "boxes" are the three value boxes (LV, MV, SV) and the three growth boxes (LG, MG, SG). These six style/cap categories were utilized in this analysis. We found that Wilshire, Russell, and Dow Jones indexes had the longest history of style-specific indexes in the Morningstar Principia database. Wilshire's six indexes (LV, LG, MV, MG, SV, SG) date back to 1978. Russell's large cap, style specific indexes (1000 Value and 1000 Growth) as well as their small cap indexes (Russell 2000 Value and Russell 2000 Growth) started in 1979. However, the Russell Midcap Growth and Midcap Value indexes didn't begin until 1986. The six Dow Jones indexes (LV, LG, MV, MG, SV, SG) began in 1980.

Barra and MSCI publish style indexes with long histories, but MSCI acquired Barra in 2004 and it is unclear which indexes will be published in the future. A number of funds track MSCI and Barra style indexes. A complete family of Barra ETFs is offered by iShares, though MSCI ETFs focus primarily on non-U.S. markets. It is unclear whether all the Barra ETFs will be offered in the future.

The November 2004 article developed and charted five portfolio management strategies using only Wilshire indexes (LV, LG, MV, MG, SV and SG). We expand this present analysis to include the Dow Jones and Russell growth indexes because there are no actionable index funds or ETFs which track the Wilshire mid cap indexes (value or growth). All the indexes (large cap, mid cap, and small cap) offered by Russell are actionable as exchange traded funds from ishares. Among the Dow Jones indexes are large cap and small cap ETF's offered by streetTRACKS.

Five different portfolio management strategies were tested over an 18 year period from 1987 to 2004. The five strategies were:

(1) Buy-and-hold each growth index (LG, MG, SG) (2) Equal annual rebalancing among the three growth indexes (3) BIPY: reallocation of all assets into last year's best performing growth index (a "momentum" strategy) (4) MIPY: reallocation of all assets into last year's middle performing growth index (a "middle path" strategy) (5) WIPY: reallocation of all assets into last year's worst performing growth index (a "bottom fishing" strategy).

Figures 1-3 show annual return data for the respective indexes color coded by the three most active portfolio management strategies: BIPY, MIPY, and WIPY. Turquoise boxes highlight the BIPY strategy, yellow boxes highlight the MIPY strategy, and the orange boxes highlight the WIPY strategy approach. The 18-year average annualized returns for the five portfolio management strategies when following their own "signal" and the Wilshire "signal" are summarized in Figure 4.

All three index families confirm that BIPY is the superior portfolio reallocation strategy when considering only U.S. growth equity indexes. Moreover, the BIPY strategy with Russell and Dow Jones indexes is improved when following the Wilshire BIPY signal.

Using Wilshire indexes, the BIPY strategy generated a 12.7% annualized return. The 18-year annualized return of the next closest growth strategies - a tie between buy-and-hold and equal annual rebalance - was 10.4% using Wilshire indexes. Among the Russell indexes following their own signal, the BIPY average annual return was 11.6%, but if Russell growth indexes are reallocated based upon the annual performance of the Wilshire growth indexes (i.e. following the Wilshire "signal") the BIPY return increased to 12.0%. Using Dow Jones growth indexes the annualized return for the BIPY strategy was 11.5%, and 12.2% if following the Wilshire signal. Pragmatically, implementing the BIPY strategy using Dow Jones indexes would require using the Dow Jones LG and SG ETFs offered by streetTRACKS and a MG index offered by ishares (either Russell-based or Morningstar-based).

Of the three index families in this study, the highest BIPY return was achieved using the Wilshire growth indexes. If the Wilshire growth indexes are set forth as providing the most reliable BIPY signal (inasmuch as the highest BIPY annualized returns were achieved using the Wilshire growth indexes), we observe that Russell missed the BIPY signal five times (1987, 1993, 1995, 1996, 2002) whereas Dow Jones indexes missed the BIPY signal six times (1992, 1994, 1996, 2001, 2002, and 2003). All three indexes agreed on the BIPY in 10 of the 18 years.

To summarize the findings to this point we present two X-Y graphs (or return/risk maps) of the five portfolio management strategies (BIPY, MIPY, WIPY, Buy and Hold, and Annual Equal Rebalancing) using Russell performance data following the Wilshire signal (see Figures 5 and 6). Both figures include growth and value indexes (where red labels represent Russell value indexes and blue labels represent Russell growth indexes).

Each data point (or "label") in Figure 5 represents the intersection of the 18-year average annualized return and the 18-year standard deviation of return. Also included in the graph are the 18-year risk/return coordinates for six Russell benchmark indexes: 1000 Growth, 1000 Value, 2000 Growth, 2000 Value, 3000 Growth and 3000 Value. The Russell 1000 index is a large cap index, 2000 a small cap index, and the 3000 a total market index.

Performance results in Figure 5 demonstrate that values indexes (both the portfolio management strategies and the raw benchmark indexes) are clearly superior to growth benchmarks and strategies over this particular 18 year period. The northwest quadrant is the most desirable space in an X-Y graph as it represents the higher return with lower volatility. MIPY (or M in the graph) was the clear winner among value indexes and BIPY was the best portfolio management strategy among growth indexes. WIPY (W in the graph) was clearly the worst strategy among value and growth indexes.

Each of the three Russell value benchmark indexes (1V, 2V, and 3V) and four of the value-based portfolio strategies (MIPY, BIPY, BH, and R) generated higher returns with lower volatility than all but one of the strategies utilizing growth indexes (BIPY). It is important to note that annual reallocation following the Wilshire BIPY signal (using three style-specific, growth-based Russell ETFs available from ishares) would have been required in only 9 of the 18 years (as noted by how many times the turquoise boxes changed columns in Figure 1). In other words, a momentum-based strategy does not imply total portfolio reallocation every year. Following the MIPY strategy (using the Wilshire signal) using three style-specific, value-based Russell ETFs (also available from ishares) would have required a reallocation of assets in 11 of the 18 years. The results up to this point have relied upon 18-year annualized returns over the period from 1987-2004. Understandably, point-to-point analysis can be particularly sensitive to the starting year and ending year of the period being studied. Our solution was to calculate rolling returns throughout the entire period to observe the performance of each strategy during smaller intervals of time. It is essentially a way to see what happened "along the way" rather than simply how things ended up. Thus, we calculated the 3-year rolling returns for the five portfolio management strategies as well as buy-and-hold results in each of the benchmark indexes (Russell 1000 Value and Growth, Russell 2000 Value and Growth, and Russell 3000 Value and Growth). There were a total of 16 three-year returns during this 18-year period. The average of the 16 three-year rolling returns and the average of the 16 three-year standard deviations of return are reported in Figure 6.

The findings in Figure 6 are startling. The various growth indexes and portfolio management strategies utilizing growth indexes show a positive correlation between risk and return whereas among value indexes the correlation is negative. For example, the BIPY strategy using growth indexes (blue labels) had the highest average 3-year rolling return, but also one of the highest levels of volatility of rolling return. Among the value indexes (in red), the highest 3-year average rolling return was achieved by the MIPY strategy. Interestingly, that same strategy had the lowest standard deviation in the 3-year rolling returns. Furthermore, all the value-based data (red labels) are in the northwest quadrant and most of the growth-based data are in or near the least desirable southeast corner of the X-Y graph.

The point of this analysis is not to advocate building a portfolio using only growth indexes, or value indexes for that matter. Rather, it suggests how the "growth portion" of an index-based portfolio might be "tactically" enhanced using a BIPY strategy. In the Editor's Corner note in the July/August 2004 issue of the Financial Analysts Journal, Robert Arnott suggests that an investor can boost return by as much as 250 basis points through systematic rebalancing, asset selection, and tactical asset allocation. In this analysis, we have shown that a BIPY strategy using growth-based indexes improved return by 180 basis points (compared to a buy-and-hold strategy or an equal annual rebalancing protocol). Using value-based indexes, our May, 2005 article demonstrated a MIPY tactical benefit of 230 basis points (compared to the next closest strategy which was annual rebalancing). Both the BIPY and MIPY portfolio management strategies involve the three elements suggested by Arnott: rebalancing, asset selection, and tactical asset allocation.

Additionally, our study has re-demonstrated the presence of a "value-premium" and how the management of value-based indexes using a MIPY strategy enhances the inherent premium in value-oriented securities. In 2001, James Davis noted that "Perhaps the biggest disappointment in the past three decades is the inability (or unwillingness) of [mutual] funds to capture the value premium that has been observed in common stock returns during the period" (Mutual Fund Performance and Manager Style, Financial Analysts Journal, Jan/Feb 2001). We propose that the MIPY strategy might be one solution to that dilemma.

Figure 1. Wilshire Growth Indexes

Figure 2. Russell Growth Indexes

Figure 3. Dow Jones Growth Indexes

Figure 4. Signal Strength

Figure 5. Average Annual Return

Figure 6. Average 3-Year Rolling Returns

____________________________________________________________________________________
Craig L. Israelsen, Ph.D. is an associate professor at Brigham Young University. He teaches family finance in the Department of Home and Family Living. His research interests include mutual fund analysis. He writes monthly for Financial Planning magazine.