Style investing: behavioral explanations of stock market anomalies

Abstract PhD-project The aim of this thesis is to explore the mechanisms of style investing. My project consists of two parts, each with an individual goal: 1. The first objective will be to analyze the implications of the dynamics of value and growth strategies for the US stock market. 2. The second objective will be to find explanations for stock returns by introducing the effects of collective preferences of investors into the dynamics of stock markets. We introduce style popularity as an important influencing factor in the investment process. By analyzing alternative methods with a more sophisticated ranking we want to show better insights in the dynamical process of value stocks. We make a distinction between switching versus fixed-style stocks. Within each style we distinguish between stocks that stay for one period and stocks that stay for two or more periods within a particular style. We analyze how stocks behave when they switch from style and what variables or factors are important to explain the style-switching behaviour. We find that only a small fraction of the value group is responsible for the value premium, namely the switching-style stocks. Theories regarding the value premium like the error-in-expectation hypothesis are tested with this new classification. This leads to new insights and conclusions for the value premium. The sub division of value and growth stocks into switching versus fixed-style stocks is critical note against style investing, because the label of value or growth stocks is limited. To profit from particular investment styles portfolio managers have to choose stocks that migrate from one style to another. The second objective is to develop an alternative perspective on the performance of style investing. Barberis and Shleifer (2003) created a model that is based on a demand-driven process. Stock returns are determined by investors who base their asset choice on a group level instead of an individual level of stocks. The investment process is in terms of investment cycles where the demand of a particular style is based on the past performance of the style. Instead of choosing a passive benchmark and trying to beat this benchmark by over- and underweighting stocks investors nowadays choose a particular style that did well in the past and hope that this will be a guarantee for future performance. We describe the popularity of investment styles as collective preferences of individuals and the changes of such preferences over time. In order to measure popularity, it will be necessary to construct a popularity index for different investment styles. Using different variables that reflect popularity we create a popularity index that reflects whether a style is popular or not. Having constructed a popularity index, it will be possible to check to what extent these portfolios are exposed to style investing or style popularity. In addition, the time series of returns form the different investment styles will be used to show that past performance is a guarantee for the attractiveness of investment styles.