The authors test their hypothesis that observable momentum in stock returns is the result of imperfect diffusion of company-specific information to the investing public. They find that except for the very smallest stocks, the potential profitability of momentum trading declines with company size. They also find that when they hold company size fixed, momentum strategies work better for stocks with low analyst coverage than for those with high analyst coverage. Finally, they find that the importance of analyst coverage is far greater for recent stock losers than it is for stock winners.
Many research studies have confirmed that stock returns show momentum over medium-term horizons of 3–12 months: Past winners continue to perform well, and poor performers continue to perform poorly.
The authors empirically investigate reasons for this phenomenon. They focus on the impact of how quickly and efficiently company-specific information is diffused to the general investing public. Their central thesis is that return momentum exists when company-specific information is not rapidly spread to investors, and they suggest that stocks with lower information diffusion exhibit greater momentum.
Because direct measurements of the efficiency of a company's information diffusion are unavailable, the authors use a company's market capitalization and the amount of its sell-side analyst coverage as primary proxy indicators. They suspect that investors are more apt to seek out information about larger companies, for which they can take larger positions; hence, diffusion will be faster for larger companies than it is for smaller companies. Furthermore, the authors suspect that the amount of analyst coverage reflects the extent that financial intermediaries are facilitating the diffusion of company-specific information.
The authors look at virtually all publicly traded U.S. stocks for the 1976–96 period and use reported monthly returns and analyst-coverage data. Extremely small, thinly traded companies are excluded, and the remaining stocks are grouped according to their periodic short-term performance. Rigorous regression and sensitivity analysis techniques are then applied, and the results confirm the main thesis: The profitability of momentum in trading declines with size, and when holding the size fixed, momentum strategies work best when analyst coverage is low. Moreover, those two factors appear to reinforce each other, because the greatest return momentum came from small stocks with low analyst coverage.
This information diffusion effect seems to be particularly pronounced for poor-performing stocks that have low analyst coverage. For these stocks, trading momentum lasts longer because the market appears to react more sluggishly to negative news on these companies. The authors explain this asymmetry by noting that a company's management is apt to aggressively release and publicize good news, whereas bad news diffuses more slowly. Therefore, short-selling recent losers may be a profitable trading strategy.