A novel sentiment measure and the stock market return

The main issue in finance nowadays is to understand the financial market dynamics. This paper attempts to shed light on the relationship between media sentiment and stock market returns. An enormous amount of data are analyzed to define media sentiment. The data are downloaded from Twitter accounts of the main relevant economic financial journals and blogs at present in Italy. After classifying each message in positive, negative and neutral through natural linguistic processing, the investor sentiment index is defined and the VAR model is adopted. The paper demonstrates sentiment index derived from social media content affects the irrational investor’s behavior, therefore the stock market return.