File Download
Supplementary

postgraduate thesis: The impact of social media sentiment on stock performance

TitleThe impact of social media sentiment on stock performance
Authors
Advisors
Advisor(s):Chau, MCL
Issue Date2017
PublisherThe University of Hong Kong (Pokfulam, Hong Kong)
Citation
Lo, K. [盧家祺]. (2017). The impact of social media sentiment on stock performance. (Thesis). University of Hong Kong, Pokfulam, Hong Kong SAR.
AbstractIncreasingly more investors are seeking information from social media to help make investment decisions. While previous research has shown that stock opinions in traditional media is a possible predictor of stock returns, no previous research has specifically considered the effects of future stock performance and moderation effect of penny stocks. In addition, there is a lack of research in applying data analytics on understanding the effect of moderators in stock opinions on future stock performance. It is also worth investigating the different effects of information in social media and traditional media for stocks with different market capitalizations and industries. The analyses were performed in three studies as follows. As information on penny stocks is often less reported in traditional media, investors may rely more on social media to obtain such information for investment advice. The first study used the net proportion of positive words in stock articles in social media to help predict the future stock performance, in particular for penny stocks in shorter terms. The moderation effect of penny stocks on the net fraction of positive words was found to be significant, revealing a stronger relationship between social media and stock performance at lower price and market capitalization levels. Consistent with the price reversion effect, penny stocks had reverse moderation effect in longer periods. Based on the findings, simple strategies utilizing social media and the measure of net fraction of positive words were proposed. Penny stocks with smaller market capitalization levels could also have higher returns compared with those with higher market capitalization levels in short terms. Although previous research has shown that sentiment analysis in social media may help predict future stock returns, the results could be inconsistent. The second study explored the effect of moderators in stock opinions on stock returns in terms of three aspects involving the quality of articles, author’s positions and social media related attributes. In terms of author’s position, it is interesting to find that the stated long and short position by the author could have a dominating effect on stock returns irrespective of the sentiment level in the related articles. Concerning social media related attributes, larger number of email subscriptions could be an important moderator in social media. Although quality of articles seems to be an important attribute, the effects were not significant in the presence of author’s positions and social media related attributes. The third study used Dow Jones news articles and social media data to explore how these two different data sources affect the future stock performance. Apart from investigating their individual and reinforcing effects, the difference in prediction performance in terms of different aspects involving the number of news articles, subset analyses on firms with different levels of market capitalizations and industries were performed. The effects for using different methods for defining the net fraction of positive words were also studied. Social media is becoming a prevalent platform for investment advice. The results for the three studies have important implications for individual investors, institutional investors and regulators.
DegreeDoctor of Philosophy
SubjectSocial media - Economic aspects
Prices - Stocks
Dept/ProgramBusiness
Persistent Identifierhttp://hdl.handle.net/10722/257616

 

DC FieldValueLanguage
dc.contributor.advisorChau, MCL-
dc.contributor.authorLo, Kar-kei-
dc.contributor.author盧家祺-
dc.date.accessioned2018-08-08T06:35:29Z-
dc.date.available2018-08-08T06:35:29Z-
dc.date.issued2017-
dc.identifier.citationLo, K. [盧家祺]. (2017). The impact of social media sentiment on stock performance. (Thesis). University of Hong Kong, Pokfulam, Hong Kong SAR.-
dc.identifier.urihttp://hdl.handle.net/10722/257616-
dc.description.abstractIncreasingly more investors are seeking information from social media to help make investment decisions. While previous research has shown that stock opinions in traditional media is a possible predictor of stock returns, no previous research has specifically considered the effects of future stock performance and moderation effect of penny stocks. In addition, there is a lack of research in applying data analytics on understanding the effect of moderators in stock opinions on future stock performance. It is also worth investigating the different effects of information in social media and traditional media for stocks with different market capitalizations and industries. The analyses were performed in three studies as follows. As information on penny stocks is often less reported in traditional media, investors may rely more on social media to obtain such information for investment advice. The first study used the net proportion of positive words in stock articles in social media to help predict the future stock performance, in particular for penny stocks in shorter terms. The moderation effect of penny stocks on the net fraction of positive words was found to be significant, revealing a stronger relationship between social media and stock performance at lower price and market capitalization levels. Consistent with the price reversion effect, penny stocks had reverse moderation effect in longer periods. Based on the findings, simple strategies utilizing social media and the measure of net fraction of positive words were proposed. Penny stocks with smaller market capitalization levels could also have higher returns compared with those with higher market capitalization levels in short terms. Although previous research has shown that sentiment analysis in social media may help predict future stock returns, the results could be inconsistent. The second study explored the effect of moderators in stock opinions on stock returns in terms of three aspects involving the quality of articles, author’s positions and social media related attributes. In terms of author’s position, it is interesting to find that the stated long and short position by the author could have a dominating effect on stock returns irrespective of the sentiment level in the related articles. Concerning social media related attributes, larger number of email subscriptions could be an important moderator in social media. Although quality of articles seems to be an important attribute, the effects were not significant in the presence of author’s positions and social media related attributes. The third study used Dow Jones news articles and social media data to explore how these two different data sources affect the future stock performance. Apart from investigating their individual and reinforcing effects, the difference in prediction performance in terms of different aspects involving the number of news articles, subset analyses on firms with different levels of market capitalizations and industries were performed. The effects for using different methods for defining the net fraction of positive words were also studied. Social media is becoming a prevalent platform for investment advice. The results for the three studies have important implications for individual investors, institutional investors and regulators.-
dc.languageeng-
dc.publisherThe University of Hong Kong (Pokfulam, Hong Kong)-
dc.relation.ispartofHKU Theses Online (HKUTO)-
dc.rightsThe author retains all proprietary rights, (such as patent rights) and the right to use in future works.-
dc.rightsThis work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.-
dc.subject.lcshSocial media - Economic aspects-
dc.subject.lcshPrices - Stocks-
dc.titleThe impact of social media sentiment on stock performance-
dc.typePG_Thesis-
dc.description.thesisnameDoctor of Philosophy-
dc.description.thesislevelDoctoral-
dc.description.thesisdisciplineBusiness-
dc.description.naturepublished_or_final_version-
dc.date.hkucongregation2017-
dc.identifier.mmsid991043976596103414-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats