Individual Stocks Have Become More Volatile

Summary of working paper 7590
Featured in print Digest

While there's no trend toward increased volatility at the market level, there is a significant trend of increasing "idiosyncratic" volatility at the individual firm level.

Is the public correct in feeling that stocks are more volatile now than ever? New research by John Campbell, Martin Lettau, Burton Malkiel, and Yexiao Xu shows that while there's no trend toward increased volatility at the market level, there is a significant trend of increasing "idiosyncratic" volatility at the individual firm level. Their findings imply that it is more important than ever to control portfolio risk by diversifying across many stocks, whether these are held directly or in mutual funds. The study also raises important questions about what exactly is driving the increase in firm-level volatility.

In Have Individual Stocks Become More Volatile? An Empirical Exploration of Idiosyncratic Risk (NBER Working Paper No. 7590), the authors calculate market, industry, and firm-level variances using daily data from July 1962 to December 1997. All firms traded on the NYSE, AMEX, and NASDAQ are included in the study (just over 2,000 firms at the beginning of the sample period, increasing to almost 9,000 at the end), but variances are value-weighted so that larger firms are given greater weight in the study. Market and industry variances have remained fairly stable over the period, but firm-level variance shows a large and significant positive trend, more than doubling from 1962 to 1997. All three forms of volatility increase substantially in economic downturns and tend to lead recessions.

The increase in firm-level volatility, with little change in market volatility, implies that correlations among stocks have tended to decrease over this period. The authors explore this directly by calculating an equal-weighted average of all the pair-wise correlations among stocks in their sample. The average correlation in the late 1990s is only about one-third the average correlation in the early 1960s. The authors argue that "declining correlations among stocks imply that the benefits of portfolio diversification have increased over time." To illustrate this, they calculate the standard deviations of portfolios containing different numbers of randomly selected stocks. The evidence reveals only a modest increase over time in the standard deviation of a typical 50-stock portfolio but a much more dramatic increase in a typical two-stock portfolio. They conclude that increasing idiosyncratic risk has raised the number of randomly selected stocks needed to achieve portfolio diversification.

The authors discuss some possible explanations for their findings. Since the 1960s the trend has been to break up conglomerates into corporations with a more concentrated focus and less diversified cash flows. Also, more companies now issue stock early in their life cycle when there is still a great deal of uncertainty about their long-run prospects. Leverage tends to increase the uncertainty of future payments to equity investors, but the authors note that this effect goes the wrong way; measured using market values, leverage has tended to decrease during the last decade which should have reduced firm-level volatility. Finally, the rise of institutional ownership may be an important influence. If institutional investors make decisions in similar ways and rely on similar information, then shocks to institutional sentiment may be an important factor driving increased firm-level volatility.

-- Anna Bernasek