If everyone followed an indexing strategy, would markets still be efficient?
This question has come up repeatedly ever since indexed strategies first appeared in the mid 1970s. Critics of indexing assert that markets would be less efficient if all investors adopted a market-fund investment approach. One can accept this theoretical viewpoint and still embrace indexing with enthusiasm.
If the adoption of indexed strategies became so pervasive that market efficiency were impaired, it would be a self-correcting process. Mispriced securities would create opportunities for investors to earn profits in excess of their research costs, and their activity would drive prices back to equilibrium levels. We will never know how much information and liquidity are required for an efficient market. Markets for consumer durables such as homes or autos appear to be at least reasonably efficient, despite very poor liquidity, high search costs, and the absence of perfectly fungible assets. This behavior suggests a shift to passive investing would have to be very pronounced to have any effect on market efficiency.
Even if all professional investment managers adopted a passive approach, other market participants would continue to provide price-setting information. Sources of such information could include corporate stock buybacks, acquisitions, and the investment activities of officers, employees, competitors, and suppliers.
Despite the impressive commercial success of indexed investing strategies over the last twenty-five years, they still represent only a fraction of total stock market wealth.
Isn't the success of indexing in recent years mostly due to a "self-fulfilling prophecy"?
Index funds appear to push up prices of a handful of big company stocks simply because they're included in the S&P 500 Index or the S&P TSX Index.
Some critics of indexing assert that mechanical buying from index funds creates a "self-reinforcing trend" in a handful of large company stocks and that their price behavior is dictated by cash inflows to index managers, not fundamental business conditions at the underlying companies.
Evidence to support this assertion is difficult to find. A more plausible explanation of pricing suggests it is the active money managers who dictate prices to indexers, not the other way around. As an example, an analysis of trading activity in General Electric Corp. stock* (the largest component of the S&P 500 index) found that programmed buying from index funds in January 1997 accounted for approximately 1.3% of total GE monthly trading volume. The notion that 1.3% of trading attributable to passive investors possessing no useful information determines the price-discovery process for the remaining 98.7% of market participants is far-fetched.
Advocates of the "self-fulfilling" viewpoint must also confront a wide disparity in performance of individual issues. If the behavior of large company stocks is primarily attributable to passive investors buying without regard to fundamental developments, it is difficult to explain why Coca-Cola shares appreciated only 0.5% in 1998 while Wal-Mart Stores soared 106.5%.
The same theory holds true in Canada. Index funds and pension funds do not "trade" a great deal in the financial sector in Canada, and yet that sector has the highest volumes of trading each and every day on the S&P TSX index. If index investing were pushing or moving the prices of the entire financial sector in Canada, then the size of each trade would be substantially larger and the positions of investments in pension funds and index funds would be constantly in flux rather than the consistency we observe.
*Strategic Insight Mutual Fund Overview February 1997
What is CRSP?
CRSP ("crisp") is an acronym for the Center for Research in Security Prices at the University of Chicago.
Established in 1960 with a grant from Merrill Lynch & Co., the center undertook a massive data-gathering project to answer the question "how have stocks performed over the long term?" Under the direction of James Lorie, Ph.D., a professor of business administration, and Lawrence Fisher, associate professor of finance, a database of both price and dividend information was compiled for all common stocks listed on the NYSE beginning in 1926.
Over two million bits of information were entered onto magnetic tape, and the commercial computers then becoming available calculated total returns. The results were published in a landmark article in the Journal of Business in January 1964. The center continues to add data on a regular basis, and with a $180,000 grant from Dimensional Fund Advisors in 1984, added data from NASDAQ markets starting in January 1972.
The CRSP data files have been an essential tool in the study of capital markets by an entire generation of financial economists.