The growth of passive investing has spurred suggestions that stock returns may be influenced by a so-called “index effect,” as trading by index funds responds to changes in index membership. If this were true, stocks added to the index would initially outperform amid buying pressure, while index deletions would underperform.
Our recent paper—“What Happened to the Index Effect: A Look at Three Decades of S&P 500® Adds and Drops”—analyzes a sample of S&P 500 additions and deletions. We focused on the S&P 500 as it is the world’s most widely followed index; USD 5.4 trillion tracked the large-cap U.S. equity benchmark at the end of 2020, up from USD 577 billion at the end of 1996. Hence, if an index effect exists anywhere, one might expect it to appear for the S&P 500.
The results from our paper corroborate the general consensus reflected in the existing literature: the S&P 500 index effect seems to be in a structural decline. Indeed, Exhibit 2 shows a clear attenuation in the magnitude of excess returns (versus the S&P 500) for sample additions and deletions between the announcement date and the effective date over time. The reduction in the index effect is independent of the source of additions, the destination of deletions, or the companies’ sector classifications.
Our paper uses an implicit cost measure to track stock liquidity, and this measure reflects the market makers who sit between buyers and sellers. Trading does not happen instantaneously; market makers may have to hold inventory from one side of a trade before the other side has completed. All else equal, the risk of holding inventory increases when the passive assets that would have to trade a constituent change represent a larger portion of a stock’s median dollar value traded (MDVT). A greater movement in a stock’s price introduces greater risk for the market maker. In each case, the market maker may pass this risk onto investors as an additional cost for providing liquidity.
Exhibit 3 shows that changes in the implicit cost measure helped to explain more than 50% of the variation in median excess returns for sample additions between 1997 and 2021. In other words, an improvement in liquidity appears to have helped explain the attenuation in the index effect.