Evening the Odds

by Craig L. Israelsen and Katie Walker
Reprinted by permission from Financial Planning Magazine, May 2006

he active vs. passive debate is often complicated by studies that distort the issues involved. We propose a three-fold approach when addressing this issue. First, studying all mutual funds may not be necessary inasmuch as the bulk of mutual fund assets are located in the largest 500 funds. Second, using the average return of several benchmark indexes is required to establish a less-biased estimate of a "passive" return. Third, measuring total assets that outperform a passive return benchmark is much more compelling than measuring the number of funds that outperform a specified set of benchmark indexes.

Issue #1: Analysis of all funds or largest 500 funds? At year-end 2005 there were just over 18,000 mutual funds and exchange traded funds (ETFs) in the Morningstar Principia database. Isolating actively managed funds that invest predominantly in U.S. equities brought the total to 5,728. (Each fund had to be classified in the "Domestic Stock" Morningstar category; each fund had to have a best-fit beta greater than or equal to 0.50; and each fund had to have at least 66 percent of its portfolio in U.S. stocks, and less than 25 percent of its portfolio in non-U.S. stocks, bonds, cash or "other." Index funds were omitted. Redundant share classes were included.) These same "filters" were applied to extract funds from the year-end Principia database for 2001, 2002, 2003, and 2004. The summary statistics are shown in "Annual Report Card".

Annual Report Card

As can be seen, the location of assets among U.S. equity funds is heavily concentrated in the largest 500 funds, which represented between 8.7 to 14.4 percent of all funds during that last five years. On average, about 78% of all U.S. equity fund assets reside in the largest 500 funds (or ETFs). In this article, we propose that studying the largest 500 funds can provide a reliable view of how the majority of the U.S. equity fund assets performed, despite that fact that 500 funds is a small fraction of the total of all funds. Specifically, we analyzed the performance of the largest 500 funds in comparison to their respective benchmark indexes. By focusing on the largest 500 funds, we effectively studied the performance of nearly 80% of all U.S. equity fund assets.

Issue #2: Should passive performance benchmarks be based on a single index or the average of several indexes? For better or worse, the performance of actively managed mutual funds and ETFs is compared to benchmark indexes, and as such, indexes become the de facto measure of passive performance. However, index makers (Russell, Wilshire, S&P, Morningstar, etc.) use unique methodologies in the construction and maintenance of indexes which causes significant variance in the performance of indexes that are ostensibly measuring the same component of the U.S. equity market (see "Variance in the Indexes"). We propose that funds should be compared to an average of several similar indexes, rather than just one index.

In 2001, for example, the average return of five prominent large growth indexes was -19.9% (green shaded data). However, the S&P 500/Citigroup Growth had a return of -12.7% while the Dow Jones Large Growth had a one-year return of -25.46% producing a staggering difference of 1,274 basis points (highlighted with blue shading). In 2003, the Wilshire Midcap Value Index had a 48.3% return, 1,310 basis points higher than the 35.2% return of the Dow Jones Midcap Value Index. In 2002, the S&P Small Cap 600/Citigroup Growth index had a return of -15.4% while the Dow Jones Small Growth posted a -38.9% return producing a difference of 2,353 basis points. The potential disparity in performance between individual indexes is larger among growth indexes and among small cap indexes.

Given the variability in performance among prominent indexes, we suggest aggregating the performance of prominent indexes (by style box) prior to engaging in comparisons of active and passively managed funds. Establishing a representative benchmark return is paramount before the "active vs. passive" analysis can be fairly conducted. Said differently, if one is attempting to assess the performance of actively managed small cap growth funds in 2002 it will make a huge difference which index is used inasmuch as the performance of the S&P Small Cap 600/Citigroup Growth Index was so different from the Dow Jones Small Growth Index. An advocate of active management would obviously choose the Dow Jones Small Growth Index, whereas the passive advocate will choose the S&P Index. We believe a reasonable solution is to aggregate index returns by style box and use averaged returns as the benchmark return against which actively managed funds are measured.

Given the variability in performance among prominent indexes, we suggest aggregating the performance of prominent indexes (by style box) prior to engaging in comparisons of active and passively managed funds. Establishing a representative benchmark return is paramount before the "active vs. passive" analysis can be fairly conducted. Said differently, if one is attempting to assess the performance of actively managed small cap growth funds in 2002 it will make a huge difference which index is used inasmuch as the performance of the S&P Small Cap 600/Citigroup Growth Index was so different from the Dow Jones Small Growth Index. An advocate of active management would obviously choose the Dow Jones Small Growth Index, whereas the passive advocate will choose the S&P Index. We believe a reasonable solution is to aggregate index returns by style box and use averaged returns as the benchmark return against which actively managed funds are measured.

Variance in the Indexes

Large Cap U.S. Equity Indexes		2001	2002	2003	2004	2005
Value Indexes	S&P 500/Citigroup Value	-11.71	-20.9	31.8	15.7	8.7
	Dow Jones Large Value	-6.11	-17.4	25.9	13.4	5.1
	Russell 1000 Value	-5.59	-15.5	30.0	16.5	7.1
	Russell Top 200 Value	-8.79	-18.0	26.8	13.3	4.6
	Wilshire Large Value	-8.17	-18.0	31.7	15.4	7.9
	Average Large Value Index Performance	-8.1	-18.0	29.2	14.9	6.7
	Basis Point Differential Between High and Low	612	534	590	315	411
Blend Indexes	Morningstar Large Cap	-15.10	-23.5	27.0	9.5	4.9
	Russell 1000	-12.45	-21.6	29.9	11.4	6.3
	Standard & Poor's 500	-11.88	-22.1	28.7	10.9	4.9
	Wilshire Large Cap 750	-12.77	-21.5	28.6	11.0	5.5
	Average Large Blend Index Performance	-13.1	-22.2	28.5	10.7	5.4
	Basis Point Differential Between High and Low	322	195	286	186	140
Growth Indexes	S&P 500/Citigroup Growth	-12.72	-23.6	25.7	6.1	1.1
	Dow Jones Large Growth	-25.46	-31.6	29.5	5.2	2.6
	Russell 1000 Growth	-20.42	-27.9	29.8	6.3	5.3
	Russell Top 200 Growth	-20.50	-28.0	26.6	3.7	2.9
	Wilshire Large Growth	-20.36	-25.0	25.4	6.5	2.9
	Average Large Growth Index Performance	-19.9	-27.2	27.4	5.6	3.0
	Basis Point Differential Between High and Low	1274	799	441	272	412
Mid Cap U.S. Equity Indexes		2001	2002	2003	2004	2005
Value Indexes	S&P Midcap 400/Citigroup Value	7.15	-10.1	40.2	18.9	10.8
	Dow Jones Midcap Value	6.87	-6.5	35.2	25.2	10.9
	Russell Midcap Value	2.34	-9.6	38.1	23.7	12.7
	Wilshire Midcap Value	7.10	-12.3	48.3	20.5	12.1
	Average Mid Value Index Performance	5.9	-9.6	40.4	22.1	11.6
	Basis Point Differential Between High and Low	481	588	1,310	628	184
Blend Indexes	Morningstar Mid Cap	-4.63	-18.1	38.4	19.7	12.7
	Russell Midcap	-5.63	-16.2	40.1	20.2	12.7
	Standard & Poor's Midcap 400	-0.60	-14.5	35.6	16.5	12.6
	Wilshire Mid Cap 500	-1.83	-17.8	43.0	18.8	11.1
	Average Mid Blend Index Performance	-3.2	-16.6	39.3	18.8	12.2
	Basis Point Differential Between High and Low	503	353	738	375	161
Growth Indexes	S&P Midcap 400/Citigroup Growth	-7.97	-19.2	31.0	14.0	14.4
	Dow Jones Midcap Growth	-17.56	-32.7	43.6	15.4	14.5
	Russell Midcap Growth	-20.16	-27.4	42.7	15.5	12.1
	Wilshire Midcap Growth	-12.72	-23.7	39.2	17.3	10.0
	Average Mid Growth Index Performance	-14.6	-25.7	39.1	15.5	12.8
	Basis Point Differential Between High and Low	1,219	1,353	1,269	333	451
Small Cap U.S. Equity Indexes		2001	2002	2003	2004	2005
Value Indexes	S&P SmallCap 600/Citigroup Value	13.10	-14.5	40.0	23.2	8.5
	Dow Jones Small Value	12.79	-2.4	43.7	18.4	6.7
	Russell 2000 Value	14.02	-11.4	46.0	22.3	4.7
	Wilshire Small Value	10.06	-13.3	51.2	20.3	8.8
	Average Small Value Index Performance	12.5	-10.4	45.2	21.0	7.2
	Basis Point Differential Between High and Low	396	1,208	1,112	486	409
Blend Indexes	Morningstar Small Cap	5.26	-20.4	47.7	20.4	5.8
	Russell 2000	2.49	-20.5	47.3	18.3	4.6
	Standard & Poor's Smallcap 600	6.53	-14.6	38.8	22.6	7.7
	Wilshire Small Cap 1750	-0.30	-20.8	47.5	18.1	7.9
	Average Small Blend Index Performance	3.5	-19.1	45.3	19.9	6.5
	Basis Point Differential Between High and Low	683	612	893	453	332
Growth Indexes	S&P SmallCap 600/Citigroup Growth	-1.19	-15.4	37.3	22.0	7.3
	Dow Jones Small Growth	-8.50	-38.9	48.5	15.5	8.8
	Russell 2000 Growth	-9.23	-30.3	48.5	14.3	4.2
	Wilshire Small Growth	-14.31	-28.9	43.9	16.1	6.8
	Average Small Growth Index Performance	-8.3	-28.4	44.6	17.0	6.8
	Basis Point Differential Between High and Low	1,312	2,353	1,121	770	463

Issue #3: Measuring outperforming actively managed assets or number of outperforming funds? A significant flaw in many active vs. passive studies is tallying the number of funds that under-perform (or out-perform) some particular benchmark index. This approach explicitly implies that all funds are of equal size, which is obviously not true. We call this incorrect approach taking a "body count". It must be agreed, however, that the ultimate measure of relevance is how the end-user-the investor-is impacted. As such, measuring the total assets that outperform this or that index is the more correct measure of "impact."

For example, a fund with $30 billion in assets that outperforms its benchmark index is more important than an "outperforming" fund with only $30 million in net assets. Likewise, consider a sample of 10 actively managed equity funds where only two outperform a specified group of benchmark indexes. If those two funds happen to have 80% of all the assets represented by the 10 funds, an asset count is a much more accurate representation of the value of active management than a body count.

As shown in "Assets vs. Body Count" 43.9% of all actively managed funds beat their respective aggregated benchmark index returns (see "Variance in the Indexes") during the five-year period from 2001-2005. Over the same time period, 52.6% of all actively managed U.S. equity assets did likewise. This difference of 870 basis points is the justification for measuring assets rather than fund count.

The data in "Assets vs. Body Count" represent the sum of annual results over the five year period. In 2001, for example, there were 613 actively managed large cap value equity funds with $413 billion in assets. Of that total, 384 funds beat the average performance of the five indexes shown in "Variance in the Indexes". Those 384 funds represented 62.6% of all the large cap value funds. However, those 384 outperforming funds had $311 billion in assets, representing 75.4% of all large cap value assets in 2001. The same calculations were applied to the funds in the remaining eight style boxes for 2001, as well as for the years 2002-2005.

Over the five year period, there were 3,667 actively managed large value funds (many of the funds survived the entire five year period and were therefore counted five times). Of that total, 1,439 beat the averaged return of the five large value equity indexes. Thus, 39.2% of all actively managed large cap value funds beat the passive return of their benchmark indexes. On the other hand, there was a summed total of $2.35 trillion of assets among the 3,667 actively managed large value funds over the five year period. Of that total, 47.5% or $1.117 trillion assets beat the averaged LV benchmark return. The figure "Assets vs. Body Count" summarizes the five year results for each style box and for the aggregate total results. As can be clearly seen, asset count vs. body count makes a significant difference in the assessment of active management.

Also shown in the same table are the results for the largest 500 U.S. equity funds over the same time period. The percentage of outperforming assets (54.5%) is very similar to the result (52.6%) for all funds over the five-year period. The "body count" differential between all U.S. equity funds and the largest 500 U.S. equity is more sizeable (43.9% vs. 48.4%). Nevertheless, the results for the largest 500 funds are very close to the results for all funds except in the small cap growth category. The similarity in results suggests that studying the behavior of the largest "x" number of funds (in this case 500) produces findings that are descriptive of what is essentially happening among all funds. Moreover, studying all funds is hardly necessary when the largest 500 funds possess nearly 80% of all the assets.

We observe that within the nine-box style grid there is considerable variation in the performance of actively managed funds relative to the performance of aggregated benchmark index returns. In the large cap growth category, about 70% of all assets outperformed the average returns of five large growth indexes over this particular five year period, whereas only about 60% of all large cap growth funds did the same. Among mid cap funds, active management appears less appealing as no more than 40% of actively managed assets or funds beat their respective benchmarks. Among small cap funds, active management generates results that beat passive performance about half of the time.

We observe that within the nine-box style grid there is considerable variation in the performance of actively managed funds relative to the performance of aggregated benchmark index returns. In the large cap growth category, about 70% of all assets outperformed the average returns of five large growth indexes over this particular five year period, whereas only about 60% of all large cap growth funds did the same. Among mid cap funds, active management appears less appealing as no more than 40% of actively managed assets or funds beat their respective benchmarks. Among small cap funds, active management generates results that beat passive performance about half of the time.

In summary, slightly more than half of all actively managed assets beat their passive benchmarks during this five year period (whether measuring all funds or the largest 500 funds), whereas slightly less than all actively managed funds did the same. The only compelling case for actively management is among large cap growth funds. However, to those who suggest that passive management is clearly superior to active management, we did not find results to support that notion. This much we know; the active vs. passive issue is a complex one and requires a level playing field to correctly assess who's beating who.

Go Toward the Light

(The summation of the percentages in each style box will not equal the “Total” as the data in

each style box are differentially weighted).

____________________________________________________________________________________
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

Katie Walker is an undergraduate student at Brigham Young University.