Hidden Measures

by Craig L. Israelsen
Reprinted from Financial Planning Magazine
March, 2006

ow did individual U.S. stocks perform compared with U.S. equity mutual funds during the past year? Assessing the performance of the U.S. equity market depends on which measuring stick you choose - and there are many. As shown in "Measuring Sticks for Stocks", the 2005 return of the Dow Jones Industrial Average (a media darling, but rarely used as a benchmark for index funds) was 1.72%, the S&P 500 returned 4.91% and the Dow Jones Wilshire 5000 (a broad equity market index) returned 6.32%. In 2005, one can say that the broader (i.e. including large, mid, and small cap firms) the scope of the index, the higher the return. This same phenomenon occurred in 2003 and 2004, whereas in 1999 the broader equity indexes (Russell 3000, Wilshire 5000) under-performed the mega large-cap indexes (DJIA, S&P 100). Hence, the choice of measuring stick has obvious consequences-you see only what it measures.

A less well-known index is the equal-weighted Wilshire 5000. Its annual performance differed significantly from other prominent U.S. equity indexes in 1999, 2001, 2002, 2003 and 2004. In 2003 the differential was particularly large. The one-year return of the other U.S. equity indexes ranged from 26.2% to 31.6%, while the equal-weighted Wilshire 5000 had a return of 91.8%. An equal-weighted index assigns equal weight to the return of each stock in the index. If you really want a challenge, start using the equal-weighted Wilshire 5000 Index as a performance benchmark. Its average annualized return (i.e. geometric mean) over the past seven years has been 21.3% compared with 2.9% for the market capitalization-weighted Wilshire 5000 and 1.7% for the S&P 500.

Measuring Sticks for Stocks

Those are some notable trends in the indexes, but how did individual stocks perform last year? In 2005, the average return of the 6,091 U.S. stocks in the Morningstar database was 7.1% (see "U.S. Stocks"), higher than all the indexes included in this study. Companies were excluded that did not have a full-year return or that were incorporated outside the U.S. The median (or middle) return for the 6,091 stocks was -1.3%, while the share-weighted return for U.S. stocks in 2005 was 2.9%. (The share-weighted return is not reported by Morningstar, but is a figure I calculated.) The share-weighted return assigns a proportionally greater weight to stocks that have more shares outstanding and less weight to thinly held stocks. It is comparable to weighting returns on the basis of market capitalization (which is calculated by multiplying a stock's current price by the number of its outstanding shares).

U.S. Stocks

Based on the 2005 returns for the seven indexes referred to earlier, as well as the mean return of all U.S. stocks, it would appear that 2005 was a mediocre year for equities, but certainly not a melt-down. Interestingly, nearly 53% of all U.S. stocks had a negative return in 2005 (data in red). In fact, among the 3,199 stocks that had a negative return, the average return was -30.6% (with a standard deviation of 25.6%). By comparison, in 2002 (a meltdown year by virtually every measure) the percentage of companies with a negative one-year return was "only" 63%. Even though 2005 was a mediocre year and 2002 was a meltdown year, the "body count" across these two particular years was surprisingly similar.

In 1999, slightly more than 54% of all stocks had a negative return, even though the U.S. equity indexes in this study all showed robust returns exceeding 20%. Moreover, the mean return for all 6,242 stocks was 42.7% and the share-weighted return for all U.S. stocks was 41.7%. However, the median return of -3.9% in 1999 revealed an underlying weakness-a weakness completely masked by the strong returns of the market-cap weighted indexes and a mean stock return which was impacted by a large number of high returns (184 firms had a one-year return of 400% or higher in 1999, with 48 of those returns in excess of 1,000%).

The performance of equities in 2005 demonstrated several similarities to 1999. The mean and share-weighted return for all stocks were both positive (though at lower levels than in 1999), the returns of prominent indexes were positive (although much lower), and yet more than 50% of all stocks had a negative return. The common denominator: a negative median return for all U.S. stocks. Though seldom reported, the median return of all stocks is a helpful measurement for two reasons: it isn't affected by market-cap weighting, and it isn't affected by outlier stocks with huge positive returns. After all, one of the rationales for broad equity indexes such as the Russell 3000 and the Wilshire 5000 is to minimize the impact of outliers.

Over the past seven years, an average of 46% of all U.S. stocks had a negative one-year return in any given year. Not surprisingly, 2002 was the worst year; nearly 63% of stocks had a negative return and the average negative return was -44.2%. In 2003, a rebound year, only 15% of all stocks suffered a negative return and the average negative return was -33.4%.

Also included in the figure "U.S. Stocks" are annual performance measurements of the 500 largest stocks each year. These stocks represent approximately 8% of all U.S. stocks (as measured by Morningstar's database), yet they account for two-thirds of all the outstanding shares from year to year. During the past seven years, an average of 45% of the largest 500 stocks had a negative one-year return each year; 2002 was the worst year where 76.4% of them had a negative return and 2003 was the best year with only 7.2% of the 500 companies experiencing a negative one-year return. A large percentage of the largest 500 stocks represent the bulk of the S&P 500 Index, by far the most "cloned" equity index on the planet. It is interesting to observe that in spite of the S&P 500's 21.0% return in 1999, over 40% of its holdings had a negative return. Market-cap weighting at work again.

What about U.S. equity mutual funds? Funds included in this study had to have a one-year return as of Dec. 31, 2005 and less than 15% of their portfolio in cash, bonds or non-U.S. stocks. Redundant funds were also omitted (i.e. only the primary share class of multiple share-class funds was included). Raw data were obtained from the January release of Morningstar Principia for each year. In other words, 1999 data were obtained from the January, 2000 release of Principia; 2000 data were obtained from the January, 2001 release, and so on. Year-end data obtained for only one year at a time as opposed to using the January, 2006 database to get data over the previous seven years minimizes the effect of survivor bias.

THE LARGE-COMPANY EFFECT

Let's connect some of the dots. In 2001, the mean return of 6,197 stocks was 18.1%, the median return was 3.5%, and the share weighted-return was -3.1%. The lower share weighted-return reveals the problem that year: bad performance among larger stocks. This is borne out by the -6.2% average return of the 500 largest stocks in 2001. As many mutual funds load up on the same large stocks, most U.S. equity funds (77.1% to be exact) lost money in 2001. Among the largest 100 funds-which represent only 4% to 5% of all mutual funds but hold about 60% of all fund assets-87% had a negative return in 2001. The cause is absolutely clear: portfolio redundancy. A lot of funds own the same large stocks. So, when large stocks do well, most funds do well.

As another example, in 1999 the average return of the 500 largest stocks was 42.2% and only 13.6% of all U.S. equity funds had a negative return. In 2003, the average return of the 500 largest stocks was 58.9% and 0.3% of all funds had a negative return. By contrast, in 2002, the average return of the 500 largest stocks was a dismal -25.0% and 97.1% of the 2,391 U.S. equity mutual funds had a negative return. That year was clearly bad for U.S. stocks, but it was a horrific year for equity funds because portfolio redundancy magnified the meltdown among mutual funds. In fact, 100% of the largest 100 U.S. equity funds had a negative return; and the average negative return was -22%.

U.S. Equity Funds

Among equity mutual funds, being "top heavy" is a double-edged sword. When a relatively small number of widely-held, large-cap stocks do well, most equity funds prosper. But, when a relatively small number of large stocks perform badly, a disproportionately large number of equity funds take a beating. Consider this: about 6% of all U.S. stocks are categorized as large caps (by Morningstar), but nearly 60% of all U.S. equity mutual funds are categorized as large cap funds. The bottom line is that there are a lot of mutual funds buying the same large-cap stocks. Practically speaking, it's therefore important to screen for portfolio redundancy when investing in more than one large-cap fund.

Finally, a few observations about expense ratios. The mean expense ratio of U.S. equity funds in 1999 was 1.23%, 1.29% in 2002, and 1.19% in 2005. Of course, these figures are equal-weighted which carries the unfortunate implication that all funds are the same size, which is clearly false. The solution is to weight each fund's expense ratio by its share of the total net assets of all funds. The result is the net asset-weighted expense ratio (see italicized data in "U.S. Equity Funds").

The net asset-weighted expense ratio for all U.S. equity funds in 1999 was 0.80%, 0.86% in 2002, and 0.76% in 2005. As the net asset-weighted expense ratio is lower than the mean expense ratio, we infer that larger funds tend to have lower expense ratios-as they should based on the notion of economies of scale.

Expense ratios (weighted by net assets) have changed over the past seven years. From 1999 to 2002, expense ratios tended to increase. After 2002, expense ratios declined-at least in the aggregate. The mean expense ratio of 1.19% in 2005 was 7.8% lower than the mean expense of 1.29% in 2002, and the net asset-weighted expense ratio of 0.76% in 2005 was 11.6% lower than the 0.86% net asset-weighted expense ratio in 2002.

It's clear that widely held large-cap stocks exert a powerful influence on the performance of the U.S. equity market and that market-cap weighting governs the typical measurement of U.S. equity performance. However, there is more to the performance of the U.S. equity market and it can only be observed through less common statistics, such as the median return and the percentage of stocks and funds with a negative return.

____________________________________________________________________________________
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