According to Standard & Poor's, the most frequently used equity benchmark today is the S&P 500. This index is widely regarded as the standard for measuring U.S. large-cap stock market performance. In total, more than $1 trillion is indexed to the S&P 500.
The S&P 500's use of a market-cap weighting makes it a good representation of the U.S economy. The index includes a representative sample of premier companies in leading industries and is used by 97% of U.S. money managers and pension plan sponsors. Unfortunately, the U.S. economy does exhibit some degree of cyclicality.
It is generally accepted that the S&P 500 is a growth-biased index. The index provider's use of a market-cap weighting creates this growth bias; it simply is a symptom of the chosen methodology. Market-cap weightings are a function of both size and relative price. More expensive companies enjoy a disproportionate weighting at the expense of cheaper value stocks. While the extreme case is a sector bubble, the relationship is always present on a relative valuation basis. This point is especially relevant with respect to reinvestment. Consider that the marginal dollar is allocated disproportionately to the most highly valued stocks at any one time.
Recent history demonstrated that speculative overvaluation increases the index's allocation to growth sectors. As growth valuations inflate, these cap-weighted problems can grow to dangerous proportions. For example, in December 1990, technology sector companies represented 6.5% of the value of the S&P 500. Ten years later, tech had grown to more than 34% of the index's composition. Clearly the upside multiple explosion affected the index's market weightings at the expense of other "value" sectors. Now, as a byproduct of the crash in technology company valuations, the tech sector accounts for just 13.6% of the index's constituency.
Impact on diversification
The most severe problem that arises with the cap-weighted index's bias is its impact on diversification. Although 500 companies are represented, 25% of the index is concentrated in just 11 positions. Ninety-percent of the portfolio is composed of just slightly more than one-half of the stated 500-stock sample.
In March 2000, near the peak of the S&P 500's valuation, the index's top four positions were Microsoft Corp., Cisco Systems Inc., General Electric Co. and Intel Corp. Combined, these four companies were worth about $2 trillion and represented approximately 16% of the market cap of the S&P 500. For every $1 billion indexed to the S&P 500 in March 2000, $160 million was directly invested in these four companies. As of September 2002, the market capitalization of Microsoft, Cisco, GE and Intel had declined by an unfathomable $1.4 trillion. Translated to comprehensible numbers, $106 million was lost for every $1 billion indexed. Also in March 2000, undiscovered fraud at WorldCom Inc., Enron Corp. and Tyco International Ltd. inflated their representation to 2.1% of the S&P 500; $18 million was lost per $1 billion indexed. Lofty prices for telecom stalwarts like Lucent Technologies Inc. and Nortel Networks Ltd. alone accounted for 3% of the S&P; $25 million was lost per $1billion indexed.
That makes nine positions considerably overvalued. These efficient "market valuations" led to an allocation of more than 20% of the S&P 500 and resulted in a loss of $150 million for every $1 billion indexed. It seems investors would not want to make such a disproportionate bet on an indexed investment.
Compounding this problem are manager assignments benchmarked to the S&P 500 and most likely selecting from the same 500 stocks. When considered within the context of total managed and indexed assets, the overlap of these positions across managers and accounts might, in fact, create an unacceptable risk.
The ideal cap-weighted structure might be to base weightings on a uniform structure like book value. Unfortunately, no such uniformity exists.
Until such time as uniformity does exist, one way to enhance the diversification of the S&P 500 would be to dilute concentrations across all 500 positions.
By swapping a portion of a portfolio's existing market-cap weighted index for an equally weighted S&P 500, institutional investors can efficiently accomplish three things. They can:
* partially hedge the growth bias of the S&P 500 by adding a neutral index allocation;
* cease systematic purchases skewed toward the highest valued securities; and
* more broadly diversify a portfolio with only 0.2% invested in each of the 500 securities in the index. (Under this methodology, the exposure to the nine stocks in the prior example would have been less than one-tenth of what they were in March 2000.)
Table 1 shows that in virtually every time period evaluated over the past 76 years, an equal-weighted index outperformed the cap-weighted S&P 500. Some claim that while this is a theoretically sound idea, transaction costs would eliminate any benefit. Recent experience, however, disproves this claim. A registered investment adviser in Chicago has operated an equal-weighted S&P 500 since 1996. Net of transaction costs, its equal-weighted index has produced 320 basis points of additional return annually over the last five years.
In the context of historical data, this would have been a very beneficial strategy for tax-preferenced investors.
Plan sponsors and others benchmarked to the S&P 500 might find it beneficial to evaluate whether adding an equal-weighted component to their index allocation could be a beneficial amendment to their index strategy.