Financial markets are doubtlessly one of the most complex structures in the known free market economy. Stock prices are impacted by an inherently complex mix of micro and macro-economic factors, henceforth the prediction of stock price movement has been a constant area of research in the industry. Hedge fund and mutual fund managers have to make decisions regarding investing in securities based on public news. In this paper, we propose an approach to quantify the relationship between sentiment of financial news and stock price movement based on multiple factors. Moreover, we use a transfer learning methodology to see if a possible correlation can exist with non-financial training data.
Authors: Mahad Afzal, Priya Yadav, Karam -El Loh, Syed, Umer Gilani