Textual sentiment analysis in international financial markets

Abstract

Textual sentiment refers to the degree of positivity or negativity in texts. Compared with investor sentiment, which has been extensively studied by behavioral finance researchers, textual sentiment includes the more objective reflection of conditions within firms, institutions and markets. This thesis extracts negative sentiment from news stories by using the dictionary-based approach (based on the Loughran and McDonald’s (2011) finance negative words), and studies its effects in both equity and debt markets.