The Dual Micro/Macro Informing Role of Social Network Sites: Can Twitter Macro Messages Help Predict Stock Prices?

Abstract

This study proposes a framework for understanding the role of Social Network Sites (SNS) in informing clients at individual message (micro) and aggregate (macro) levels. To validate the micro/macro informer framework, we examine if an aggregate of Twitter messages can be used as a predictor of future stock prices of publicly traded companies. Our field study analyzes Twitter posts related to 18 Fortune500 companies using Latent Semantic Analysis to extract the semantic and conceptual content in the form of key themes. Using the factors comprising the themes, we fit a regression model that uses tweet volume and tweet topic strength to predict 8.3% of variability in stock prices beyond what is explained by the fluctuations of the stock market. Our results suggest that Twitter can be viewed as a macro informer for stock markets. This confirms our conjecture that SNS are both platforms for micro informing and macro informers. Our study should stimulate informing science researchers to examine other cases of macro informing and the role of data mining and text mining technologies in message aggregation for macro informing.

Combining structure, content and meaning in online social networks: The analysis of public's early reaction in social media to newly launched movies

“…In backchanneling a medium sustains an online conversation about an event of interest. Examples include monitoring audience of TV programs and new movies release (Asur and Huberman, 2010;Nielsen, 2014;Highfield and Bruns, 2013;Wakamiya et al, 2011) or understanding and predicting traders' reactions in the stock market events (Evangelopoulos et al, 2012). Of particular interest for this work is the role of backchannel environments in collaborative meaning-making (McNely, 2009) and as a perspective to design tools for the visualization of large-scale ongoing conversations on Twitter (Dork, Gruen, Williamson, & Carpendale, 2010).…”

Extracting and evaluating conversational patterns in social media: A socio-semantic analysis of customers’ reactions to the launch of new products using Twitter streams

“…The Twitter data flow generated by specific events through backchanneling provides a more focused opportunity for social media mining. Examples include monitoring TV audience reactions (Harrington, Highfield, & Bruns, 2013;Nielsen, 2014;Wakamiya, Lee, & Sumiya, 2011), understanding ad predicting traders' reactions in the stock market (Evangelopoulos, Magro, & Sidorova, 2012), or measure the diffusion of ideas in online discussions (Tsur & Rappoport, 2012). In all these cases the objective is to extract meaningful and representative feedback from online conversations and exploit this feedback to help decision makers to identify effective ways to influence attitudes and behavior.…”

Predicting the Use of Twitter in Developing Countries: Integrating Innovation Attributes, Uses and Gratifications, and Trust Approaches

“…Twitter stimulates users to engage in timely and up-to-date activities with families, friends, and co-workers, sharing information, news, and opinions (Honeycutt & Herring, 2009). Therefore, the Twitter platform serves as an infrastructure in which, on the one hand, individual informers deliver messages to individual clients and, on the other hand, because of the large public nature of the information exchanged, Twitter itself becomes the informer, providing aggregate information to clients (Evangelopoulos et al, 2012).…”

“…New communication tools have become a major issue within society and have captured the attention of researchers around the world as a significant research topic (Park & Kim, 2013). SNSs have led to the publication of a number of studies in different contexts such as the stock market, parliamentary and presidential elections, emergency events, crises, health, education, business, and so on (Evangelopoulos, Magro, & Sidorova, 2012;Lacka & Chong, 2015;Liu, Liu, & Li, 2012;Pere-Latre, Portilla, & Blanco, 2010;Prybutok, 2013). Owing to the differences in SNSs' pur-…”