Essays on stock splits and herding

Abstract

This thesis consists in an analysis of stock splits, and their relationship with dispersion of beliefs and herding. Chapter 1 introduces the topics that I tackle throughout the thesis. In particular, I motivate the interest in herding and stock splits presenting the unifying interpretation line among each chapter. Chapter 2 proposes a literature review on stock splits, focusing on the explanations that the theoretical literature suggests and the empirical evidence of the market reaction. Chapter 3 reports the results of an empirical analysis around the time of a stock split on the relation between the dispersion of beliefs among investors and the market reaction and future performance of the splitting company. We provide empirical results on a sample of US splits which occurred from 1993 to 2004. They show that, at the time around the announcement of a split, the distribution of the analysts forecasts changes in mean and dispersion. Moreover, an event study shows that the di¤erences of opinion have an impact on the future performance of the splitting firms and on the motivations behind the event. Chapter 4 focuses on a literature review of herding, and in particular on the empirical investigation of imitative behavior among institu- tional investors. Chapter 5 examines the relation between herding and stock splits. By herding we mean the abnormal correlation of trades among insti- tutional investors, according to the methodology developed by Sias (2004). We use data on the buying and selling activity of US insti- tutional investors, from 1994 to 2005. The results show a significant level of convergence in the overall market, both for splitting and non- splitting companies. We decompose this effect into the contributions of several types of herding. We observe the significant impact of informational cascades on the splitting stocks sample, while reputational herding and characteristic preference have a relevant impact on the non-splitting sample. The evidence of informational content in the split event is confirmed by the stabilizing effect of herding we find in the future returns of splitting companies. Chapter 6 concludes, summarizing the main results and contributions of the thesis.