Measuring the Impact of Individual Twitter Accounts on the S&P 500 and Investment Strategies

Samer Ajour El Zein, Michelle Fleming

With the stock market seeing some of its most volatile days in history, it is more important than ever for investors to be able to reduce and mitigate risk for their investments (Froot, Scharfstein & Stein, 1993). There is the constant push for higher returns and more accurate predictions from investors as well as from higher up executives on analysts and portfolio managers in order to reach target earnings. Due to these increased pressures, investors may feel the need to take on more risk in their portfolios than they normally would in order to achieve their benchmarks. The need for better foresight is always a sought after attribute for those looking to invest in portfolios or individual companies. From previous research, it has been alluded that the social media platform, Twitter, can be used as a short term stock market indicator with relative accuracy (Rao & Srivastava, 2012). It has been recorded that the social media platform had an accuracy rate of nearly 90 percent from a 2011 study looking at the daily up and down changes for the Dow Jones closing values (Mao & Zheng, 2011). While in the past, studies have been focused on the social media platform as a whole, our study looked at the link between the accounts of influential individuals and movements in the market. The results of this paper have come to the conclusions that personal Twitter accounts can have the ability to be used as a stock market indicator for the S&P 500 as well as a way to craft better performing portfolios. From the results, there is a link between observed variables relying heavily on the concept of sentiment analysis and retweet count per trading day. With an accuracy rating of 91.2 percent in the predictions based on the Ordinary Least Squared (OLS) Model and the ARIMA time series analysis model, there are enormous opportunities for the financial industry to implement these findings in order to enable smarter and less risky investments. With future developments in enabling Twitter extraction, we believe that individual twitter accounts can be used in order to craft better performing portfolios in the market, mitigating risk and outperforming conventional methods.

Volume 12 | Issue 1

Pages: 459-484

DOI: 10.5373/JARDCS/V12I1/20202649