Research gaps on investigating sentiment extraction from stock blogs and its effect on stock direction and volume

Author: Iloka Benneth  Chiemelie

Published: 15-October-2014

On a general note, the internet has emerged into a source of aggregate vital information for stock investors in terms of decision making. This is because it has effectively changed how information is delivered to investors as well as the way these investors can act based on the information they have (Barber & Odean, 2001). From that view, the internet has been able to alter the way people invest, trade, acquire and distribute information (Zhang & Swanson, 2010). From the initial stages, it was used primarily for aggregating public information like public news, financial data and updates in the market, but the internet now covers areas such as social media and WEB 2.0 (Ullrich et al., 2008); effectively offering access to user generate contents (UGC) that can incorporate both private and public information (Tumarkin & Whitelaw, 2001). Observations have also revealed how such Vital Investing Communities (VIC) lie the Raging Bull and Yahoo finance have been influence the decision of investors with their publication of important and valuable UGC data such as proprietary analysis and recommendations on investments. Through these channels, UGC is able to enrich investors with needed information for better decision making and also allow investors to monitor the thought and process of other people’s decisions. As such, it is very important for both researchers and practitioners to gain an understanding of how individuals interact with each other in virtual communities and how such interaction influences their decision making process.

Indeed, the area of sentiment analysis, its relationship with consumers’ decision making and how it influences the overall movement of stocks is an active area of research that have produced numerous streams of interesting literature. One of such involves interaction established between communities that focuses on topics like network externalities, homophily and reputation (Chen et al., 2009; Gu et al., 2007; Zhang, 2009). Other research have been done in the area of understanding how activities in the virtual communities are correlated with the market outcome investigation predictors in terms of volume, disagreement, and bullishness of such positing s (Antweiler & Frank, 2004; Das & Chen, 2007; Das et al., 2005; Sabherwal et al., 2008).

Notwithstanding such studies, there are a number of gaps in this area of research. Irrespective of the high belief that sentiments from VICs can actually be used to predict stock value, little evidence have been published in order to back that believe (Tumarkin & Whitelaw, 2001; Das & Chen, 2007; Antweiler & Frank, 2004). From general point of view, it is believed that when people engage in discussions, they exchange vital information to each and enlighten others based on their personal view and experience. If such is the case, the believe moves on with the idea that the informed (mostly those that are not familiar with the area of such information) can make use of the information they have acquired in decision making. While this is possible, it neglects the fact that information can be false, reducing the potential of such information having any impact when adopted.

Another research gap is the classification of sentiments as either optional/objective or factual/subject. While it is slightly reflected in the first gap, the understanding presented here is that while majority of the researches done in sentiments analysis have grouped the effects as either being positive or negative [Pang, Lee & Vaithyanathan 2002; Turney 2002], there is still a wide gap in terms of analyzing sentiments based on the extent of their originality as either being a fact or being purely an individual opinion [Wiebe, Wilson & Bell 2001; Wiebe, Wilson, Bruce, Bell & Martin 2004]. When in subjective nature, the effects of sentiment analysis can be negative because sentiments are based on individual view instead of facts related to the stock being analyzed, thus investors can actually make wrong decisions.

The third gap to be considered in this research is the extent of influence that sentiment analysis can have on level of stock movement. Although evidence exist in terms of the belief that sentiment analysis can influence stock movement, there are little or no evidence to validate the extent of such influence on stock movement e [Wiebe, Wilson & Bell 2001; Wiebe, Wilson, Bruce, Bell & Martin 2004]. The importance of such research is that it will help to understand whether the influence are negligible or highly significant.

From the above analysis, a number of gaps exist in terms of understanding how sentiment analysis can influence the performance and movement of stocks across both lines. Such gap present a major need to explore these areas and expand existing understandings as well as create new understandings on how sentiment analysis influence the performance and movement of socks.

References

Antweiler, W. & Frank, M. (2004) “Is all that talk just noise? The information content of internet stock message boards.” Journal of Finance. 59(3):1259-1295.

Barber, B. & Odean, T. (2001) “The Internet and the Investor.” Journal of Economic Perspective 15:41-54.

Chen, M., Gu, B. & Konana, P. (2009) “Social Capital, Social Identity and Homophily Behavior in Virtual Communities: An Analysis of User Interactions in Stock Message Boards” Revise and Resubmit Information Systems Research.

Das, S. & Chen, M. (2007) “Yahoo! For Amazon: Sentiment Extraction from Small Talk on the Web” Management Science 53(9):1375-1388.

Das, S., Martinez-Jerez, A. & Tufano, P. (2005) “eInformation: A Clinical Study of Investor Discussion and Sentiment” Financial Management 34(3):103-137.

Gu, B., Konana, P., Rajagopalan, B. & Chen, M. (2007) “Competition Among Virtual Communities and User Valuation: The Case of Investing-Related Communities” Information Systems Research 18(1):68-85.

Pang, B., L. Lee and S. Vaithyanathan 2002. Thumbs up?: Sentiment classification using machine learning techniques. Association for Computational Linguistics, Philadelphia, PA.

Sabherwal, S., Sarkar, S. & Zhang, Y. (2008) “Online talk: does it matter?” Managerial Finance 34(6):423-436.

Tumarkin, R. & Whitelaw, R. (2001) “News or noise? Internet postings and stock prices” Journal of Financial Analyst 57(3) 41-51.

Tumarkin, R. & Whitelaw, R. (2001) “News or noise? Internet postings and stock prices” Journal of Financial Analyst 57(3) 41-51.

Turney, P. D. 2002. Thumbs Up or Thumbs Down? Semantic Orientation Applied to Unsupervised Classification of Reviews. Association for Computational Linguistics (ACL'02), Philadelphia, PA.

Ullrich, C., Borau, K., Luo, H., Tan, X., Shen, L., Shen, R. (2008) “Why web 2.0 is good for learning and for research: principles and prototypes” In Proceedings of the 17th International Conference on World Wide Web, 705-714.

Wiebe, J., T. Wilson and M. Bell 2001. Identifying Collocations for Recognizing Opinions. Association for Computational Linguistics, Toulouse, France.

Wiebe, J., T. Wilson, R. Bruce, M. Bell and M. Martin 2004. Learning Subjective Language. Computational Linguistics 30(3): 277-308.

Zhang, Y. & Swanson, P. (2010) “Are day traders bias free? – evidence from internet stock message boards” Journal of Economics Finance 34:96-112.

Zhang, Y. (2009) “Determinants of Poster Reputation on Internet Stock Message Boards” American Journal of Economics and Business Administration 1(2):114-121.