Do Investors Under-React to ROE?
72 Pages Posted: 1 Feb 2008
Date Written: January 31, 2008
Previous researches document a delayed market response to public information (Bernard, and Thomas 1989, 1990; Abarbanell and Bernard 1992; Daniel, Hirshleifer and Subrahmanyam 1998; Foster, Olsen and Shevlin 1984). In this study, we uncovered the investors' under-reaction to ROE (Return on Equity), commonly known as the cross sectional gauge of firms' profit generating efficiency. Using in excess of 31 years' of firms' fundamental financial information and stock return data, we investigate, employing three asset-pricing models including Fama and French 3-factor model (1993), Fama-French 4-factor model and a proprietary Jensen Measure 4-factor model, the stock return performance on deciles of portfolios formed based on holding firms' ROE ranking. We found that, over time, all three models are able to explain a significant portion of the variations in returns of stocks ranked on ROE while the proprietary Jensen Measure 4-factor model provides the best explanation. A quarterly rebalanced portfolio holding the top 10 percent highest ROE stocks will earn an annual abnormal excess return of approximately 15.2 percent due to investors' under-reaction. We also discovered that investors' under-reaction to ROE diminishes with decreasing ROE level and almost vanishes at the lowest two deciles of firms ranked by ROE level. More importantly, investors have been persistently under-reacting to ROE information since 1973, especially the recent period from 1980 to 2004 during which the annual abnormal excess returns stayed at over 16 percent.
Keywords: Return on Equity, Under-Reaction, Fama-French Three Factor, Asset Pricing, Momentum
JEL Classification: G14
Suggested Citation: Suggested Citation