Scholarlink

<p>Social media have attracted a lot of attention in the financial market. Opinions on social media can serve as an indicator of future stock performance. However, prior literature shows inconsistent results when using sentiment analysis (e.g., positive and negative words) on social media to predict future stock performance. The inconsistent results suggest that omitted-variable bias may exist in the relationship between social media elements and future stock performance. To investigate this issue, this study tests the effect of <i>author attributes</i> on using social media to predict future stock performance. Using data collected from a popular online stock opinions forum, our results show that an author’s stated long and short positions in social media articles have a significant predictive power on future stock performance. Meanwhile, stated long position and author’s popularity partially moderate the predictive power of sentiment level on future stock return. Furthermore, we investigate whether investors’ reactions vary across market contexts by assessing the boundary condition of platform trust. Our study helps explain the inconsistent results reported in the literature. This study also provides new insights into author attributes on social media and the role they play in the financial market, as well as provides investment and management directions for individual investors, fund managers, firms, and regulators.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Whose Investment Advice Do You Follow? Studying the Effect of Author Attributes on Using Social Media Stock Opinions To Predict Stock Performance

  • 期刊论文

  • 发表日期：

    2025年11月15日

查看全文
  • Michael Chau,
  • Bingjie Deng,
  • Kar Kei Lo,
  • Dongjun Wei,
  • Wanghongyu Wei

摘要

Social media have attracted a lot of attention in the financial market. Opinions on social media can serve as an indicator of future stock performance. However, prior literature shows inconsistent results when using sentiment analysis (e.g., positive and negative words) on social media to predict future stock performance. The inconsistent results suggest that omitted-variable bias may exist in the relationship between social media elements and future stock performance. To investigate this issue, this study tests the effect of author attributes on using social media to predict future stock performance. Using data collected from a popular online stock opinions forum, our results show that an author’s stated long and short positions in social media articles have a significant predictive power on future stock performance. Meanwhile, stated long position and author’s popularity partially moderate the predictive power of sentiment level on future stock return. Furthermore, we investigate whether investors’ reactions vary across market contexts by assessing the boundary condition of platform trust. Our study helps explain the inconsistent results reported in the literature. This study also provides new insights into author attributes on social media and the role they play in the financial market, as well as provides investment and management directions for individual investors, fund managers, firms, and regulators.