Institutions and Individuals at the Turn-of-the-Year




    Search for more papers by this author
    • Sias is from Washington State University, and Starks is from the University of Texas. We thank Keith Brown, Amar Gande, Mark Griffiths, Jeff Heisler, Ed Maberly, René Stulz, participants at seminars at Indiana University, the University of Missouri, the University of Texas-Dallas, Tulane University, and Washington State University, and an anonymous referee. We thank the NYSE for providing the Torq data and Joel Hasbrouck and George Sofianos for their assistance in interpreting the Torq data. We thank Vera Tonry for research assistance.


This article evaluates the tax-loss-selling hypothesis against the window-dressing hypothesis as explanations for turn-of-the-year anomalies. We examine differences between securities dominated by individual investors versus those dominated by institutional investors and find that the effect is more pervasive in the former. Controlling for capitalization, we find that in early January (late December), stocks with greater individual investor interest outperform (underperform) stocks with greater institutional investor interest. These results hold for both stocks that previously appreciated in value and stocks that previously depreciated in value. The results are most consistent with the tax-loss-selling hypothesis as an explanation for the turn-of-the-year effect.