Date of Graduation
The existence of market anomalies has long been recognized in the finance literature. Several studies have documented the effects of size, dividend yields, E/P ratios, book-to-market value ratios, turn of the year (January effect), investor overreaction, and mean reversion on market returns. Still, much controversy surrounds the existence of, and explanations for the observed market anomalies. In this dissertation, I attempt to overcome such difficulties in analyzing the existence of the market anomalies of investor overreaction and mean reversion, while providing recommendations for future research. In the first chapter of the dissertation I briefly discuss past research in the areas of investor overreaction and mean reversion, while motivating the need to explore alternative methods of examining these issues. Chapter 2 of the dissertation examines perceptions of the investor overreaction anomaly to changes in various methodological issues. I examine the sensitivity of the results from past research to changes in the method of calculating returns, portfolio formation periods, failure to properly account for delisting firms, as well as the source of data. The results tend to reject the overreaction hypothesis, suggesting that investors do not overreact to recent information while ignoring long-term trends. Chapter 3 of the dissertation investigates the problems associated with examining mean reversion using portfolios formed on the basis of market capitalization and standard econometric techniques. Chapter 3 provides an alternative method of analyzing mean reversion as a result of such problems. The results from this chapter clearly indicate that portfolio prices predictably revert toward their fundamental or trend value following a one-time shock. Additionally, by exploiting such information, it is possible to devise an investment strategy that outperforms other commonly used investment strategies. Chapter 4 of the dissertation analyzes mean reversion by applying the techniques used in Chapter 3 to portfolios containing stocks that are classified by various industry groups. The results confirm those from the previous chapter, implying that asset prices do not follow an integrated process. Chapter 5 provides a summary of the dissertation and discusses areas of future research related to the behavior of asset prices.
Gropp, Jeffrey Michael, "Mean reversion in United States stock prices: A panel approach." (2000). Graduate Theses, Dissertations, and Problem Reports. 8961.