Overreaction of Dow stocks

Abstract

Several studies have found mean reversion in monthly stock returns over long horizons. However, these studies can be challenged for several reasons, including the neglect or possible misspecification of risk premia. The current paper analyzes daily Dow returns over short horizons, which obviates the most serious issues in long-horizon studies using monthly data. There is strong evidence of overreaction in that large positive and (especially) negative returns tend to be followed by persistent, substantial, and statistically persuasive reversals over the next 10 days.

Keywords:

• stock price overreaction
• regression to the mean

Public Interest Statement

We often overreact to new information because we neglect the role of chance. For example, we misinterpret variations in test scores as changes in ability rather than fluctuation in scores about ability. In the same way, investors often overreact to good or bad news about a company’s fortunes, causing excessive fluctuations in stock prices which are partly reversed as time passes.

Additional information

Notes on contributors

Gary Smith

Gary Smith is Fletcher Jones Professor of Economics at Pomona College. His research encompasses financial markets and statistical reasoning, particularly stock market anomalies and the abuse of data. Several papers debunk fanciful claims, such as Asian-Americans are susceptible to heart attacks on the fourth day of the month; people with positive initials (like ACE) live several years longer than do people with negative initials (like ASS); and people can postpone death until after the celebration of meaningful events, like their birthday, Passover, and the Harvest Moon Festival. He has also explored regression to the mean in a variety of contexts, including testing, sports, business, forecasting, and investing—including the results reported here.