Brexit has us all on Edge

Abstract

In this piece of research, we examine the evolution of sentiment proxies, commonly used to account for investor behaviour in the financial economics literature. Advancements in computing techniques, such as sentiment analysis and natural language processing (NLP), have allowed the creation of investor sentiment proxies directly from textual news data. Recent literature identifies that such proxies may be used to predict movements in financial assets, particularly during heightened periods of investor sensitivity - recessions, bad news cycles, etc. Following the techniques employed in other markets, we construct a sentiment indicator for the Irish Stock Market, and evaluate it against returns for the ISEQ 20 Index. With Brexit introducing uncertainty into the Irish market as a whole, we look to expand the construction of sentiment proxies from one news source, as per the existing literature, to many. Following this we evaluate the predictive power of the indicator created.