Smith Faculty Opinion Article

March, 2007

By Dr. John A. Haslem, Professor Emeritus
 E-MAIL WEB SITE

 Haslem

S&P 500 Index Mutual Funds
Diverse expenses and performance characteristics

by John A. Haslem, H. Kent Baker and David M. Smith

The commodity-like nature of index mutual fund portfolios suggests that price competition should be more evident than with actively managed funds. After all, fund managers operate index funds not to beat their benchmarks or actively managed funds, but to mimic benchmark portfolios and performance, less expenses. Therefore, fund expenses should be singularly important in explaining and predicting differences in index fund performance. Haslem (2003) treats the broader issues of fund expenses.

With vigorous price competition, asset size and 12b-1 fees should account for most of the differences in expenses for index mutual funds with the same benchmarks. However, as discussed by Ptak (2004), the realities are quite different:

Index funds are a logical starting point . . . given their commodity like nature (which should, in theory at least, breed fierce price competition). Index funds that track the same benchmark, such as the S&P 500 index, are virtually identical in terms of holdings. They also cost very little to operate since management merely entails purchasing the securities in the same proportion as the bogy and periodically rebalancing in response to asset flows or reconstitution of the benchmark. In other words, differentiation is nonexistent, barriers to entry are modest, and, it would seem, cost is king.

However, you wouldn't know it judging from the diversity of fees levied on index funds tracking the same benchmark. Elton, Gruber and Busse (2004) find that the reality underlying the market for S&P 500 index mutual funds is explained by the presence of uninformed investors and fund distributors with economic incentives to sell inferior funds. Thus, we should not be surprised (but perhaps dismayed) by the presence of large differences in management fees and expense ratios.

As to the implications of findings of large differences in management fees and expense ratios, Ptak (2004) concludes that:

The stakes in the cost battle are enormous. While low costs greatly improve a funds chance of outperforming peers while also moderating risk, the same high costs that have hampered fund performance have been an absolute boon to the fund industry.

The U.S. Securities and Exchange Commission (2000) has also voiced serious concern about the size of mutual fund expenses. The SECs expressed interest is not from the standpoint of regulating fund expenses, which optimally should be market determined, but from its role in regulating fund directors, who approve fund expenses.

Mission

This study's mission is to provide investors, researchers, and regulators with: (1) an improved understanding of the dispersion of management fees and expense ratios among the mutual funds that track the S&P 500 index; and, (2) the association of fund performance measures and characteristics to classes of both dispersed management fees and expense ratios. These measures and characteristics include the Sharpe ratio, Jensen's alpha, Morningstar Star Rating, annualized total return (3-yr.), portfolio turnover, 12b-1 fees, and average net assets.

Sample And Method

The initial sample of 202 investor- and institutional-class S&P 500 index mutual funds is from Morningstar (2005), and includes single share class funds and each class of multiple share class funds. Thus, each share class is a fund, and the number of unique portfolios is fewer than the number of funds. S&P 500 index mutual funds represent 41 percent of the total of 498 index funds. In addition, there are 83 distinct S&P 500 portfolios, which represent 35 percent of the total of 239 distinct portfolios. However, due to incomplete data, it was necessary to reduce the sample to 171.

The method in Haslem, Smith and Baker (2005) is used to identify the S&P 500 index mutual funds with (1) above (below) average and (2) varying degrees of statistically high (low) management fees and expense ratios. Following this analysis, the selected performance and related characteristics are related to each of the statistical classes of fund costs.

The methodology used is a simple, probabilistic application of standard deviations. The standard deviation provides an objective way to classify mutual funds. This quantitative measure enjoys wide acceptance.

The distribution-free Chebyshev's inequality is also applied because there is no certainty that a normal distribution applies for the financial variables under consideration. DeFusco et al. (2004) and Bottomley (1999) note that the likelihood of observing a management fee or expense ratio two or three standard deviations above the mean is relatively small, even if the variable is not normally distributed.

First, the index mutual funds with each of (1) management fees and (2) expense ratios that exceed the means of funds that track the S&P 500 index are classified by the significance of these cost dispersions using standard deviations. Means and standard deviations for each of management fees and expense ratios are calculated for institutional-class and investor-class funds. This process ensures that each funds fees and expenses are classified using statistics derived from peer funds. After all funds are placed in standard deviation classes for each of management fees and expense ratios (relative to their peer means), the institutional-class and investor-class samples are re-aggregated for further analysis.

The first standard deviation class for each of management fees and expense ratios contains index funds with costs within one standard deviation above the sample means. The other three standard deviation classes for each of management fees and expense ratios identify funds with costs that exceed the means of all S&P 500 index funds by one, two, and three standard deviations, respectively.

The costs of index mutual funds with each of management fees and expense ratios in each of the increasing cost standard deviation classes are labeled as (1) above average, (2) high, (3) very high, and (4) extremely high, respectively.

Second, index mutual funds with each of (1) management fees and (2) expense ratios that are below the means of all funds that track the S&P 500 index are also classified by the significance of these cost dispersions using standard deviations.

Following the above method, the first standard deviation class contains S&P 500 index funds with each of management fees and expense ratios that are within one standard deviation below the means. The other standard deviation class identifies funds with each of management fees and expense ratios that are smaller than the means of all S&P 500 funds by one standard deviation.

The costs of index mutual funds with each of management fees and expense ratios in each of the decreasing cost standard deviation classes are labeled as (1) below average, and (2) low, respectively. Third, the means of portfolio characteristics (performance measures and related variable) are computed and evaluated for the S&P 500 mutual funds with each of management fees and expense ratios in each standard deviation class of high and low-cost funds. Diversity Of Management Fees And Expense Ratios Figures 1 and 2 provide selected parameters of the sample of index mutual funds that track the S&P 500 index. Figure 1 identifies the sample mean management fees and expense ratios as 0.17 percent and 0.65 percent, respectively. The standard deviations are 0.14 percent and 0.46 percent, respectively. Note that the standard deviation of management fees is very large, which also reflects differences in the ways fund advisers account for management fees.

Figure 2 indicates that the range of the sample mean cost measures in the six standard deviation classes is very large. Management fees range from nearly zero percent in the 1 standard deviation class to 0.80 percent in the +3 class. Expense ratios range from 0.19 percent in the 1 standard deviation class to 2.71 percent in the +3 class. This finding, as much as any, indicates just how much S&P 500 mutual funds are not priced as real commodities.

Portfolio Characteristics: Management Fees

First, the mean performance and related characteristics of S&P 500 index mutual funds with statistically high management fees are presented. With one exception, 12b-1 fees, the single largest or smallest mean characteristic is found in the +2 standard deviation class. The 12b-1 has none of its characteristic means in the +1, +2 and +3 standard deviation classes.

Management Fees

(1) Sharpe Ratio smallest mean (0.04) is in the +2 standard deviation class;

(2) Jensen's Alpha smallest mean (-1.53) is in the +2 standard deviation class;

(3) Morningstar Star Rating smallest mean (2.60) is in the +2 standard deviation class;

(4) Annualized Total Return (3-yr.)smallest mean (2.00 percent) is in the +2 standard deviation class;

(5) Portfolio Turnover largest mean (168.83 percent) is in the +2 standard deviation class;

(6) 12b-1 Fees none of the three largest means are in the +1, +2 or +3 standard deviation classes; and

(7) Average Net Assets ($mm)smallest mean ($407.85) is in the +2 standard deviation class.

Next, information is presented on the Pearson correlation coefficients for S&P 500 mutual fund characteristics versus management fee standard deviation classes. The correlation analysis includes all standard deviation classes, from low through extremely high fees. The Sharpe ratio, Jensen's alpha, and annualized total return are negatively correlated with management fee standard deviation classes. These findings are expected, and the correlation coefficients are statistically significant at the 0.01 level. But while both portfolio turnover and 12b-1 fees are positively associated with management fee standard deviation classes, the relationship is significant only for turnover (0.01 level). These findings are expected. Morningstar Star ratings are not significantly correlated (0.10 level) with management fee standard deviation classes, but this is not totally unexpected, given prior research on this measure. Finally, average net assets are not statistically correlated (0.10 level) to management fee standard deviation classes. A negative correlation is desirable, but the lack of shared economies is not unexpected.

Portfolio Characteristics: Expense Ratios

The mean portfolio characteristics of S&P 500 mutual funds with expense ratios are presented below. With one exception (portfolio turnover), all characteristics have their three largest or smallest means in the +1, +2 and +3 standard deviation classes.

Expense Ratios

(1) Sharpe Ratio three smallest means (0.06, 0.06 and 0.09) are in the +1, +2 and +3 standard deviation classes;

(2) Jensen's Alpha three smallest means (1.19, 1.25 and 3.45) are in the +1, +2 and +3 standard deviation classes;

(3) Morningstar Star Rating three smallest means (2.62, 2.75, 2.00) are in the +1, +2 and +3 standard deviation classes;

(4) Annualized Total Return (3-yr.)three smallest means (2.35, 2.29 and 0.01) are in the +1, +2 and +3 standard deviation classes;

(5) Portfolio Turnover largest mean percent (490) is in the +3 standard deviation class;

(6) 12b-1 Fees three largest means (0.66, 0.67 and 1.00) are in the +1, +2 and +3 standard deviation classes;

(7) Average Net Assets ($MM)three smallest means ($155, $78 and $11) are in the +1, +2 and +3 standard deviation classes.

Next, information is presented on the Pearson correlation coefficients for S&P 500 index mutual fund characteristics versus expense ratio standard deviation classes. The correlation analysis includes all standard deviation classes, from low through extremely high expense ratios. The Sharpe ratio, Jensen's alpha, annualized total return, and Morningstar Star ratings all are negatively correlated with expense ratio standard deviation classes. The correlation coefficients are statistically significant at the 0.01 level. Thus, S&P 500 funds with low performance measures are strongly related to statistically high expense ratios. These findings are expected. P o r t f o l i o turnover and 12b-1 fees are both positively associated with index mutual fund expense ratio standard deviation classes, at the 0.01 level. These findings are expected, especially for 12b-1 fees, which are a component of the expense ratio. Finally, average net assets are significantly negatively correlated (0.01 level) with expense ratio classes. This well-known finding is expected.

Index Funds With Low Expense Ratios

Twenty-five (15 percent) of the 171 S&P 500 Index mutual funds have expense ratios in the 1 (low) standard deviation class. These few statistically low-cost investor (Inv) and institutional (Inst) class funds include:

(1) Barclays Global Investors S&P 500, Inv (WFSPX);

(2) California Investment S&P 500, Inv (SPFIX);

(3) Evergreen Market Index, Inst (F00APK);

(4) Fidelity Spartan 500 Index, Inv (FSMKX);

(5) Gartmore S&P 500 Loc, Inv (GRMLX);

(6) Mutual of America Equity Index, Inst (F001I5);

(7) Schwab S&P 500 eShares, Inv (SWPEX);

(8) Schwab S&P 500 Select, Inv (SWPPX);

(9) Scudder Equity 500 Index, Inst (BTIIX);

(10) Scudder Equity 500 Index, Inv (BTIEX);

(11) SSgA S&P 500 Index, Inv (SVSPX);

(12) State Street Equity 500 Index A, Inv (STFAX);

(13) State Street Equity 500 Index Svc, Inv (STBIX);

(14) T. Rowe Price Equity Index 500, Inv (PREIX);

(15) TIAA-CREF Institutional S&P 500 Index (TISPX);

(16) Transamerica Index, Inv (TPIIX);

(17) United Association S&P 500 Index, Inst (UASPX);

(18) United Association S&P 500 Index, Inv (UAIIX);

(19) USAA S&P 500 Index Member, Inv (USSPX);

(20) Vanguard 500 Index Admiral, Inv (VFIAX);

(21) Vanguard 500 Index, Inv (VFINX);

(22) Vanguard Institutional Index (VINIX);

(23) Vanguard Index Institutional Pl (VIIIX);

(24) Vantagepoint 500 Stock II, Inv (VPSKX); and

(25) Waterhouse 500 Index, Inv (TDSPX).

Vanguard leads the way with four index mutual funds with statistically low expense ratios. The total of only 25 statistically low-cost index funds is not good news for investors, however, especially as institutional funds are included. Further, there are no index funds with statistically very low or extremely low expense ratios.

Index Funds With Very High And Extremely High Expense Ratios

Five (3 percent) of the 171 S&P 500 Index mutual funds have expense ratios in the +2 (very high) and +3 (excessively high) standard deviation classes. These index funds with statistically very high or extremely high expense ratios include:

  • Expense Ratios (+2 standard deviation class)

  • MassMutual Select Index Equity A, Inv (MIEAX);

  • MassMutual Select Index Equity N, Inv (MMINX);

  • Scudder S&P 500 Stock C, Inv (KSACX); and

  • Scudder S&P 500 Stock B, Inv (KSABX).

  • Expense Ratios (+3 standard deviation class)

  • ProFunds Bull Svc, Inv (BLPSX).

The news is better here, with only five S&P 500 mutual funds with statistically very high and extremely high expense ratios. And, not unexpectedly, all five of these highcost funds are investor (Inv) class funds. Scudder and MassMutual each have two index funds with statistically very high expense ratios.

Conclusions

For investor and institutional class index mutual funds that track the S&P 500 Index, there are just 25 funds with statistically low expense ratios (management fee findings are found above). However, there are only five index funds all investor class with statistically very high and extremely high expense ratios. Thus, these results contain both bad and good news for investors.

Unfortunately, the complete story of high expense ratios for S&P 500 index funds finds more bad news. The Sharpe ratio, Jensen's alpha, annualized total return, Morningstar Star ratings, and average net assets are all statistically negatively correlated with index funds with statistically high expense ratios. Further, portfolio turnover and 12b-1 fees are statistically positively correlated with index funds with statistically high expense ratios.

References

Bottomley, H., Chebyshev's Inequality and a One-Tailed Version, 1999, http://www.btinternet.com/~se16/hgb/cheb.htm.
DeFusco, Richard A., McLeavey, Dennis W., Pinto, Jerald E. and David E. Runkle, Quantitative Methods for Investment Analysis, 2nd ed., Charlottesville, Va.: CFA Institute, 2004, pp. 137-138.
Elton, Edwin J., Gruber, Martin J. and Jeffrey A. Busse, Are Investors Rational? Choices Among Index Funds, Journal of Finance 59(1), 2004, pp. 261-288.
Haslem, John A., Mutual Funds: Risk and Performance Analysis for Decision Making. Oxford: Blackwell Publishing, 2003.
Haslem, John A., David M. Smith and H. Kent Baker, Identification and Performance of Equity Mutual Funds with High Management Fees and Expense Ratios, Journal of Investing 15 (Spring), 2007, forthcoming.
Morningstar, Principia Pro for Mutual Funds Advanced Module, January 1, 2005 (CD).
Ptak, Jeffrey, SEC May Put the Squeeze on Pricey Index Funds, Morningstar.com, February 9, 2004.
U.S. Securities and Exchange Commission, Report on Mutual Fund Fees and Expenses. Washington, D.C: Division of Investment Management, December 2000.

This article is reprinted with the permission of the Journal of Indexes, which is the publication of record for the index industry. Any rebroadcast or distribution of this content requires the expressed permission of the Journal of Indexes. All Journal of Indexes content, including past columns by Professor Haslem can be accessed on www.indexuniverse.com/JOI.

John A. Haslem, Professor Emeritus of Finance, University of Maryland.
H. Kent Baker, University Professor of Finance, American University, Washington, DC.
David M. Smith, Associate Professor of Finance, University at Albany, SUNY. Professor Baker is a doctoral graduate of the Smith School and Professor Haslem was his dissertation advisor.