Comparing equity mutual fund returns to an index can be very deceiving.
Given the performance differentials that can exist among major U.S. equity indexes, it is prudent to calculate the average return of several indexes before making comparisons against individual funds or equity portfolios. The five providers of equity indexes evaluated here include Barra, Dow Jones, Morningstar, Russell and Wilshire.
In Figures 1 through 3, the one-, three-, five- and ten-year performance figures for various U.S. equity indexes (large-cap, mid-cap, small-cap) are listed. They are further broken out according to the Morningstar style box, i.e. value, blend and growth. More importantly, the average (or aggregated) return for each group of indexes is also provided.
The "problem" of variance among equity indexes is most evident among small-cap indexes. For instance, the three-year performance figures for small-cap growth indexes ranged from 2.60% to 13.69%. Similar differentials, though often not as large, are observed among mid-cap and large-cap indexes-particularly among growth indexes.
As can be clearly seen, the choice of which index is selected as the "benchmark" against which a particular equity mutual fund will be measured is a crucial decision. Very simply, indexes in the same style box are not identical. The solution to variance among equity indexes is to calculate a group average and then use that average as the benchmark performance figure for that particular style box.
Having determined the average performance of major equity indexes in each style box, the next logical step is to determine which U.S. equity funds are meeting or exceeding the group average. Despite much rhetoric regarding the superiority of passive index-based funds versus actively managed funds, there are many actively managed funds which provide significant and consistent "excess return," or return in excess of the return of benchmark equity indexes.
This analysis seeks out such funds using a sensible approach. First, "index" performance is determined by the average return of four to five indexes (i.e. benchmarks), not just one index. Second, a modified information ratio is used to screen funds that provide returns in excess of aggregated benchmark index returns, while also taking into account volatility of return. Third, modified information ratios are calculated for the most recent 12, 36 and 60 months and then summed using a weighted formula. Fourth, only currently available funds are considered. It's largely pointless to highlight funds that are closed or available to only a subset of the general population.
Figures 4 through 6 present the top ten performing funds in each Morningstar style box as of November 30, 2004. The following filters were used as initial screens in the Morningstar Principia database: U.S. equity funds with at least five years of performance history; no purchase constraints; initial investment requirement of $25,000 or less; and at least 50% of the portfolio is in U.S. equities. Requiring a ten-year performance history would have been preferred, but too many funds were eliminated with that requirement.
Redundant share classes were also removed, meaning that only one share class of a particular fund remained in the final data set. For example, Merrill Natural Resource A (in the large growth style box) has three share classes (A, B, and C). Only share class A is listed in the top ten funds even though each share class likely qualified. This filtering technique prevents funds with multiple share classes from being over-represented.
Funds that made the top ten list had the highest modified information (IR) ratios for the previous one-, three- and five- year periods. The five-year IR was weighted more heavily than the one-year and three-year IR. Load-adjusted returns were used in calculating the IR so as to prevent load funds from being unfairly advantaged. Volatility of return is an inherent component of the IR calculation; thus funds with high volatility can only rise to the top of the list by having a correspondingly high "excess return."