But what bears even more traits of active management is the fact that these are only guidelines -- which the S&P Index Committee can, and often does, disregard. For example, turnover of the S&P's components soared in the 1990s. Only seven stocks were replaced in 1992, compared with 57 in 2000.
One tech stock after another was added to the index, and several met neither the profitability requirements nor the seasoning requirements. Of course, these stocks didn't last long on the index and were removed when their market values collapsed. More recently, the index has latched onto one of the greatest real estate booms in American history; the committee started adding real estate investment trusts to the index for the first time in late 2001. There are now 14 REITs in the index. This is all reminiscent of the kind of "style drift" for which active mutual fund managers are often derided. Evidence of this tendency in the S&P 500 can be seen by measuring changes in the index's divisor. The divisor keeps the index level stable during constituent changes. Since the S&P 500 is market-cap weighted, unless the replacement company was of exactly the same size as the company being deleted, the index level would change overnight were it not for adjustments to the divisor. It is typical for the divisor to rise a percentage point or two over the course of the year, as new entrants often rank around the middle of the S&P 500 while those removed often rank at the bottom of the index in terms of size. However, in the late 1990s, this relationship fell apart. Whether consciously or unconsciously, the S&P Index Committee skewed the S&P 500 heavily in favor of mega-cap growth stocks, often tech-related, in turn causing the index divisor to soar to more than four times its average (see chart below). We now know, of course, that this shift in favor of mega-cap tech stocks turned out to be a bad call, but the bottom line is that it was, in fact, a judgment call. In other words, active management.| Signs of Active Management Changes in S&P 500 divisor go against the market-cap index principle |
| Source: Standard & Poor's |



