MIPY Rides Again

Craig L. Israelsen and Patrick McDonough
Reprinted from Financial Planning Magazine
June 2007

he asset allocation approach we call MIPY was introduced in the November 2004 issue of Financial Planning and expanded in later Financial Planning articles, most recently in April 2006. MIPY requires annual reallocation into the middle performing index from the prior year among a specific portfolio of indexes. This article updates the April 2006 analysis, which illustrated MIPY using a portfolio of seven indexes. The period analyzed is the 25-year period from 1982-2006. Raw performance data were obtained from Morningstar Principia.

During 1982-2006, MIPY outperformed the annually rebalanced (AR) equally-weighted seven index portfolio by 178 basis points and a buy-and-hold (B&H) seven index portfolio by 265 basis points over the 25-year period. At a comparable risk level, a 40% MIPY portfolio (explained in detail later) outperformed AR by 70 basis points. The returns discussed in this article are pre-tax and exclude commissions. Annual allocation should of course take advantage of capital gains tax rates, or preferably take place within a tax-sheltered account. As always, past performance is no guarantee of future results.

Our objective in this article is to illustrate simple approaches to annual rebalancing using a realistic diversified portfolio of actionable index funds and/or exchange traded funds. We propose that seven indexes are sufficient to do this. The seven indexes in our model portfolio are: Russell 1000 Value Index (R1000V), the Standard & Poor's Midcap 400 Index (S&P400), Russell 2000 Value Index (R2000V), Wilshire REIT Index (REIT), Morgan Stanley Capital International Europe, Australasia, Far East Index (EAFE), Morgan Stanley Capital International Emerging Markets Index (EM), and Lehman Brothers Aggregate Bond Index (BOND).

This portfolio provides broad asset exposure while setting a formidable challenge for any tactical asset allocation approach. These indexes are mostly weakly to moderately correlated with one another and are all available as index funds or exchange traded funds ("ETFs"). Over the 25 year period of our analysis, there was a 72% chance of at least five of the seven indexes having positive returns and an 84% chance of at least four indexes having positive returns in any given year.

All other things equal, weakly correlated indexes are better portfolio choices. Among these seven indexes there were a total of 21 possible pair wise correlations. Over this particular 25-year period, 9 of the correlations were weak (.31 or lower) and 8 were moderate (between .35 to .59). All the U.S. equity indexes are weakly or moderately correlated both with foreign equity indexes and with the LB index. The strongest four pairwise correlations were among the U.S. equity indexes. The highest 25-year correlation (of .87) was between the Russell 1000 Value Index and the S&P Midcap 400 Index. Correlations among indexes do change slightly from year to year. Recent work both by academics and investment managers has noted that the correlations between U.S. and foreign equities have changed over time, but it is unclear whether there is a definite direction toward higher or lower correlations.

The Russell 1000 Value Index includes the Russell Midcap Value Index. Thus, we selected the S&P MidCap 400 inasmuch as it offers a blend of growth and value stocks. Moreover, it offers a different mid-cap value exposure due to the fact that Russell and S&P use different methodologies in their construction of indexes. The Wilshire REIT Index adds additional value orientation with only relatively weak or moderate correlation with U.S. large caps, foreign and bond indexes. Among fixed income indexes, the intermediate-term Lehman Aggregate Bond index is a commonly used proxy for the broad, investment-grade U.S. market. The Lehman 1-3 Yr Government Index is also a good choice, but had a lower return than the Aggregate Bond Index.

The MSCI EAFE covers almost all of the market which is non-U.S. developed countries. EAFE is the most popular foreign developed country equity index. MSCI Emerging Markets (EM) covers most of the equity markets in developing countries. EM improves foreign returns and has mostly weak correlation with other indexes. As the MSCI EM Index started in 1988, we used annual returns from the MSCI Pacific ex. Japan Index from 1981-1987 and then the EM Index from 1988-2005. The Pacific ex. Japan Index has a .85 correlation to the EM Index and a .74 correlation to the EAFE Index over the past 10 years.

Using this seven asset portfolio, we tested four annual reallocation protocols over the 25-year period from 1982-2005 against a baseline B&H approach. The buy-and-hold approach consisted of investing equally in each of the seven indexes and making no reallocations over the 25-year period. The four portfolio reallocation protocols were:
(1) Equal-weighted, annually rebalancing at the start of each year (AR). The annual return of each asset was weighted at .1429 (or 1/7). The weighted returns were then summed to determine the portfolio return each year.
(2) All portfolio assets were reallocated into the prior year's middle performing index at the start of each new year. This approach is referred to as MIPY (Middle Index in Prior Year).
(3) All portfolio assets were reallocated into the prior year's best performing index at the start of each new year. This approach is referred to as BIPY (Best Index in Prior Year).
(4) All portfolio assets were reallocated into the prior year's worst performing index at the start of each new year. You guessed it, this approach is referred to as WIPY (Worst Index in Prior Year).

By referring to "The Portfolio Players", you can see how this works by following the color coding. The 1981 annual return for each asset was needed to initiate the MIPY, BIPY, and WIPY protocols beginning in 1982. Among the seven assets, the bond index had the middle return in 1981, therefore the color of bond return in 1982 is yellow signaling that all portfolio assets were invested in the bond index at the start of 1982 based upon its middle performing return during the prior year. Blue coding refers to the BIPY protocol and the pumpkin color denotes the pattern of WIPY-based returns. In summary, the color of each box represents its return in the prior year. If the box has no color, it was not used in the MIPY, BIPY, of WIPY protocol in that particular year.

At the beginning of 2007 (for example), all portfolio assets would have been reallocated into the Russell 2000 Value Index based upon the MIPY protocol, because its return of 23.48% in 2006 was the middle performing return. The BIPY reallocation protocol would have required all assets to be invested into the Wilshire REIT Index. The WIPY approach would utilize the Aggregate Bond Index in 2007. Clearly, the MIPY, BIPY and WIPY protocols are extremely aggressive in that they require complete portfolio reallocation into one index each year.

MIPY Wins Again

If we measure the differences in portfolio performance using averaged three-year rolling returns over the 25-year period, MIPY outperformed B&H by 226 basis points, AR by 173 basis points, BIPY by 174 basis points, and WIPY by 833 basis points (see the lower portion of "Rolling Along").

We also simulated a series of hybrid approaches (40% of portfolio reallocated according to the MIPY, BIPY, or WIPY protocol with the remaining 60% split equally among the other six indexes). As shown in "Beyond Buy-and-Hold", the annualized return of the 40% MIPY protocol was only 108 basis points below 100% MIPY, but the risk (i.e. standard deviation of return) was reduced by 31% (or 618 basis points). In addition, 40% MIPY outperformed 40% BIPY and 40% WIPY. The 40% MIPY protocol outperformed AR by 70 bps and B&H by 157 bps over the 25-year period.

As shown in "Rolling Along", the primary impact of these hybrid approaches (40% MIPY, etc.) is significantly better risk-adjusted return (as measured by "Return/Std Dev" and "Sharpe Ratio"). When measuring performance using the average of 23 three-year rolling returns, a 40% annual allocation to the MIPY index and a 10% allocation to each of the other six indexes outperformed B&H by 117 basis points and AR by 64 basis points, with 69 basis points greater volatility than AR but 327 basis points less volatility than B&H. The 40% MIPY approach was also superior to 40% BIPY and 40% WIPY protocols. BIPY underperformed AR by 48 basis points using annualized returns and one basis point using averaged three-year rolling returns-thus providing a reminder that chasing returns doesn't usually turn out well. And WIPY was a bad idea no matter how you measure it (annually or three-year rolling returns). Thus, the "dogs" theory doesn't hold up well in our analysis.

We have found that BIPY tends to perform better with a portfolio of growth-oriented indexes, which are more volatile than value indexes (see our July 2005 Financial Planning article). WIPY was by far the worst approach. WIPY's poor performance is consistent with the findings of academic studies. Investors and advisors sometimes confuse WIPY-type approaches with mean reversion. The WIPY approach bets that last year's loser will be this year's winner. However, long-term mean reversion cycles are typically on the order of four to eight years, not one or two years. So WIPY likely has nothing in common with a mean reversion approach.

The rest of our analysis in this article focuses on three of our asset allocation approaches: AR, MIPY, and 40% MIPY; as the other approaches (using the indexes we have selected for this study) fall short of these three on the basis of risk-adjusted performance. But first, a thought from John Bogle.

"We must base our asset allocation not on the probabilities of choosing the right allocation, but on the consequences of choosing the wrong allocation." This April 6, 2000 quote, attributed to John Bogle, founder of Vanguard, raises several questions. So, we ask, how does MIPY and 40% MIPY compare to annual rebalancing (AR)?

During 16 of the 25 years, the annual returns of a MIPY approach exceeded those of AR by an average of 1,103 basis points. However, there were 9 years in which MIPY under-performed AR by an average of 1,209 basis points. Annual returns with a 40% MIPY protocol exceeded those of AR in 16 years by an average of 331 basis points. In the nine remaining years, 40% MIPY underperformed AR by an average of 363 basis points.

For the 23 three-year rolling periods, MIPY outperformed AR in 15 periods by an average of 515 basis points. Again, this is partly offset by MIPY underperforming AR by an average of 479 basis points in the remaining eight periods. During the same 23 three-year rolling periods, 40% MIPY outperformed AR by an average of 165 basis points in 15 periods, and underperformed AR by an average of 133 basis points in the remaining 8 periods. Thus, MIPY and 40% MIPY outperformed AR almost two-thirds of the time, whether in single years or in three-year rolling periods.

Responding to John Bogle's concern from the beginning of this section, both MIPY and 40% MIPY hold up well against AR on a year-to-year comparison, over three-year rolling periods, and over the most recent three, five, and ten year periods.

The 14.58% annualized return of the 40% MIPY protocol over the 25-year period was 70 basis points better than the 13.88% return of AR. We highlight this comparison inasmuch as AR is a well-recognized portfolio management technique. It is important to note that this is 70 risk-adjusted basis points of improvement. As shown in "Rolling Along" the two measures of risk-adjusted return ("Return/Std Dev" and "Modified Sharpe Ratio") are nearly identical using annualized returns for AR and 40% MIPY and comparable if using averaged three-year rolling returns. Moreover, 70 bps points translated to nearly $43,000 additional dollars over the 25 years period.

Some may ask "why not simply invest solely in the total stock market index?" The Wilshire 5000 Index (best represented by the Vanguard Total Stock Index fund) averaged a healthy 12.95% between 1982 and 2006. Nevertheless, AR among these seven indexes beat the total stock index by 93 basis points and with 20% less volatility (12.48% standard deviation for AR compared to 15.65% for Wilshire 5000). Very simply, a non-cap-weighted portfolio of seven indexes clearly adds value over and above a single, cap-weighted mega index. As already shown, a 40% MIPY approach was an improvement on AR. In fact, from 1982-2006, a $10,000 dollar investment in the total stock index grew to $209,969 (before taxes and inflation). The 40% MIPY approach turned $10,000 into just over $300,000. Interestingly, the total stock index and 40% MIPY both had negative annual returns 20% of the time over the 25-year period, but the average negative annual return was -4.2% for 40% MIPY compared to -9.8% in the total stock index. Reducing the severity of the "down years" is what keeps clients in the saddle.

We have reviewed hundreds of finance journal articles on asset allocation and excess returns. Some researchers find strong evidence for excess returns over limited time periods using customized portfolios and reallocating monthly or even daily while others do not. The title of a 2004 Journal of Portfolio Management article by Burton Malkiel sums up the skeptics' view: "Can Predictable Patterns In Market Returns Be Exploited Using Real Money? Not Likely."

In any case, outperforming a benchmark means doing 15% better, not 50% better. In his Editor's Corner article 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 the three steps of systematic rebalancing, asset selection, and tactical asset allocation.

The 40% MIPY portfolio management approach includes Arnott's three steps and captures almost 30% (70 basis points) of his hypothesized 250 basis point boost in return over a 25-year period. This annual approach is very simple and provides diversification over the U. S. equity and bond markets and most of the international equity markets.

Users Manual

To quote Pippen from The Lord of the Rings, "The closer we are to danger, the further we are from harm". We're uncertain if Pippen was considering a 100% MIPY approach when he said that, but it's something to ponder. In the meantime, consider a 40% MIPY protocol.

The Portfolio Players

Blue shading = best return in prior year

Yellow shading = middle return in prior year

Pumpkin shading = worst return in prior year

Beyond Buy-and-Hold

Rolling Along

____________________________________________________________________________________
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. Learn more about Craig Israelsen at http://familyliving.familylife.byu.edu/faculty/israelsen.htm

_______________________________________________________________________
Patrick McDonough is a chief financial officer for high-tech startups and lives in California.