An empirical investigation of the effect of investor sentiment, corporate governance and 52-week high on momentum returns

Abstract

This thesis investigates the existence of momentum returns in recent years and the effect of different explanatory variables on momentum returns in the pursuit of finding the source of momentum. It first tackles the existence of momentum in recent years by showing that momentum returns are large and significant in the last decade, contrary to what recent work have claimed that momentum has disappeared since the 1990s. The thesis also studies for the first time the effect of corporate governance of firms – measured using the G-index – on momentum returns. Results show that investors are more (less) reluctant to buy (sell) winners (losers) stocks that have dictatorship corporate governance structure when further good (bad) news arrives in the market, leading to mispricing (significant negative/positive alphas). Such results indicate that corporate governance influences the decision making of investors when buying and selling stocks. The thesis also studies the effect of multiple variables that, whilst having an established relationship with momentum returns individually, have yet to be tested as variables simultaneously. The disposition effect, anchoring effect and cognitive dissonance are simultaneously tested on momentum returns, to examine whether they affect momentum returns to the same degree. Results show that investors who primarily focus on a stocks’ position relative to its 52-week high will make a positive momentum return due to winners in pessimistic periods and due to losers in optimistic periods. The findings are novel and show that anchoring reinforces the disposition effect, while investor sentiment has the ability to override anchoring and to either weaken or strengthen the disposition effect. This study gives a better understanding of momentum returns, anchoring and the role of investor sentiment in terms of predicting future return continuations or reversals.