Print
E-MailRandall Morck is Jarislowsky Distinguished Chair and Distinguished University Professor at the University of Alberta, and M. Deniz Yavuz is Hanna Rising Star Associate Professor at Purdue University Krannert School of Management. This post is based on their NBER working paper. Related research from the Program on Corporate Governance includes Index Funds and the Future of Corporate Governance: Theory, Evidence, and Policy by Lucian Bebchuk and Scott Hirst (discussed on the Forum here and here); The Specter of the Giant Three by Lucian Bebchuk and Scott Hirst (discussed on the Forum here); and The Limits of Portfolio Primacy (discussed on the Forum here) by Roberto Tallarita.
Savings increasingly flow to low-cost index funds, which simply buy and hold the stocks in a major index, such as the S&P 500. This study presents direct evidence that increased indexing impairs the flow of information into stock prices. Specifically, similar idiosyncratic foreign currency shocks move the idiosyncratic stock returns of firms sensitive to those currencies 60% less when the firm is in the S&P 500 index than in proximate times when it is not in the index. We rely on currency shocks only because they are relatively simple to measure and quantify, so our findings suggest information flow impairment in general.
Malkiel (1973), arguing the stock market sets stock prices to fundamental values so well that picking stocks is pointless, proposed a radical new investment strategy: passive (buy-and-hold) investment via index funds. Since Vanguard launched the first index fund in 1973, assets under passive management worldwide have grown to US$14.5 trillion in 2022. This is perhaps the most consequential contribution of academic research to the investment management sector.
However, no innovator escapes rebuke. Malkiel’s critics charge indexing with undermining the very stock market informational efficiency the strategy relies on. As Robert Shiller explains in a 2017 interview, “indexing … is really free-riding on other people’s work … So people say, ‘I’m not going to try to beat the market. The market is all-knowing.’ But how in the world can the market be all-knowing, if nobody is trying — well, not as many people — are trying to beat it?”
Exploring whether increased indexing impedes information from entering stock prices requires distinguishing information-driven from behavioral stock price movements. To do this, we rely on Paul Samuelson’s dictum, quoted in Shiller (2001, p. 243): “markets show considerable micro efficiency. … [but] considerable macro inefficiency.” This is because stock market index returns exhibit excess volatility (Shiller 1981) and behaviorally-driven stock market manias and panics (Kindleberger 1978), whereas event studies linking news events to idiosyncratic (stock minus to market) returns “work” and are a standard methodology in many fields. We therefore devise an empirical methodology to see if specific information events induce smaller idiosyncratic stock price changes when a stock is in an index versus in nearby times when it is not.
Prior work on indexing and stock market information efficiency has mixed conclusions, perhaps reflecting different proxies used for informational efficiency. Some researchers use firm-specific stock return volatility, which theoretical work links to less information entering stock prices (Jin and Myers 2006, Veldkamp 2006; Wei and Zhang 2006). However, Roll (1988) allows that firm-specific stock price movements might also reflect “investor frenzy” and some studies thus deem firm-specific variation noise. Another class of proxies include deviations from random walks to proxy for informational inefficiency; however indexing might alter the prominence of various deviations from a random walk without changing overall informational efficiency (Baltussen et al. 2019). Encompassing all these concerns is the possibility of endogenous changes in the nature, magnitude, visibility, frequency and characteristics of shocks that a firm experiences when in versus not in an index.
Our approach is designed to overcome these problems. First, rather than relying on a proxy for informational efficiency, we directly measure how much a given information shock moves stock returns. Second, we use idiosyncratic (unrelated to US market returns) components of foreign currency shocks, which are determined in global markets, unlikely to be affected by any single firm being added to or dropped from the index, and thus unlikely to be endogenous as discussed above. In addition, currency shocks are readily observable by all market participants, and this observability remains unaffected by any changes in the information environment of any given firm. We consider and reject the alternative hypotheses that firms hedge currency risk more intensively or have less currency-sensitive fundamentals when in the index.
We find firms’ stocks to have an economically and statistically significant 60% lower idiosyncratic sensitivity to idiosyncratic foreign currency shocks when the firms are in versus not in the S&P 500, whereas the magnitude of the firm-specific currency shocks does not change significantly. This effect grows larger over time as the importance of indexing rises.
Our findings appear to extend to other firm-specific shocks. We observe a significant reduction in the standard deviation of daily firm-specific returns when stocks are part of the index relative to when they are not, particularly after 1990s, sample period characterized by increased popularity in index investing.
These results are consistent with indexing impairing the incorporation of firm-specific information into stock prices and validate Shiller’s concerns about indexing, a finding of considerable public policy importance. Escalating indexing renders share price changes less informative, less useful in providing feedback about corporate decisions (Bond et al. 2012), and thereby leaves corporate resource allocation less efficient (Wurgler 2000; Durnev et al. 2004; Chen et al. 2007; Morck et al. 2013), which has negative implications for overall economic prosperity. Increased passive investment having such effects implies that active investment has important positive externalities. We welcome further research into these issues.
