Second, professional money managers as a group are limited in their ability to compensate for this type of bias. Technically, they have the ability to eliminate the gap between erroneous expectations and true value via arbitrage (transaction strategies designed to profit from fluctuations in market value). But in practice, the use of arbitrage by professional managers is constrained by what economists call agency issues: investment styles, concerns about deviations from benchmark performance, and institutional practices.
To be sure, the distortion of any one company’s share price diminishes over time; perception eventually catches up with reality. But the transition can take years. Behavioral finance has recently gained increased legitimacy because it has helped investors realize significant returns by exploiting the gap between the irrationally perceived value of shares and their actual potential value.
Past Dogs, Future Performers
Perhaps the most common misperception that leads investors and corporate decision makers to prefer glamour over value assets is the simple effect of hindsight. We saw this in our study of U.S. stock market performance. We employed a common research methodology: sorting stocks into monthly portfolios, which were then sorted into deciles as measured by market-to-book value, looking at the years 1975 through 2004. (See "Methodology.") We then calculated the returns to investors over a five-year horizon for each decile. The decile results were aggregated for each of the monthly portfolios. We characterized the top decile as “market stars” (our own name for glamour stocks) and the bottom decile as “market dogs” (equivalent to value stocks).
The sample includes all U.S. stocks that had publicly traded equity between 1975 and 2004 (for all or part of the time
This process was repeated each month from January 1975 to December 1999, and the statistics presented in the exhibits are time-series averages of the monthly portfolio returns over the sample period. Balance sheet and operating income data is matched with stock price data to reflect publicly available information at the time of portfolio formation. The source for all data is the Compustat database from Standard & Poor’s.
Just as behavioral economists might predict, the market dogs consistently and substantially outperformed the market stars in shareholder value creation. Indeed, over the period of our research, the dogs outperformed the stars by an average of nearly 20 percent annually. Market dogs exceeded the market returns by 13 percent annually. (See Exhibit 1.) Other researchers have found results consistent with our data for nearly all equity markets in Europe and Asia.
We investigated the data set to determine what operating factors investors could have used to distinguish market dogs from market stars. One factor provides a clear correlation. As Exhibit 2 shows, investors use past operating performance to characterize companies.
We then compared actual past operating performance with the future expectations required to justify the current stock price. As expected, there is high positive correlation between past performance and future expectations. However, past performance is negatively correlated with actual performance. (See Exhibit 3.) Established valuation methodology prescribes that stocks be valued on future (not past) earning streams. Yet, many investors simply — and erroneously — assume that a company growing earnings at 10 percent per year will continue to do so. This extrapolation, unsupported by actual market history, is the primary driver of the difference in returns between the market dogs and the market stars.