Enhancing Financial Decision-Making Using Social Behavior Modeling
Abstract
Financial trading is a social activity that involves every participant's decision making. Meanwhile, people's online behavior collectively creates the public emotion which affects investors' reactions and hence market movements. This process can be modeled by connecting online social behavior and future trading behavior to better understand mechanisms of the stock movement so as to assist financial decision making. In this paper, we investigate the query information of financially related Wikipedia pages, and show that early signs of trading volume movements can be detected which expose financial risks. We embed this information into a classic pairs trading strategy acting on a large portfolio of stocks. Over 23% profits are seen when testing on the year of 2013 and 20% comes from the inclusion of online social data.