Can Real-Name Verification Enhance Stock Forums Information Quality?
In the realm of financial markets, social media stands out as a dynamic force propelled by retail investors, serving as a nexus of collective wisdom. Stock forums not only disseminate valuable insights but also facilitate the seamless flow of information, offering a diverse and unbiased array of information, and enriching the financial landscape with a comprehensive range of perspectives.
The significance of social media as a pivotal information source in the financial sector is underscored by the recognition garnered through the Best Paper award at the 2024 CAPANA Research Annual Conference. Their research, which utilized Datago's "Guba Analytics for China Retail Investor Sentiment" and "Xueqiu Analytics for China Retail Investor Sentiment", examines the impact of real-name registration implementation on social media platform information quality.
User Anonymity and Information Quality of Social Media: Evidence from a Natural Experiment
Author: Kanyuan Huang¹, Yakun Wang¹, T.J. Wong², Tianyu Zhang¹
¹ The Chinese University of Hong Kong, Shenzhen, ² University of Southern California
Abstract
We examine how removing user anonymity affects the quality of information on investment-related social media platforms. Using a difference-in-differences research design that exploits the differential timing of adopting real-name verification policies by the two most popular investment-related platforms in China, we find that the information quality of the treatment platform significantly improves after the implementation of the policy. This result is primarily driven by users who post more actively in the pre-period. These users become less active in the post-period but the content of their postings become more informative, indicating that the verification policy encourages them to be more prudent in expressing opinions. Additionally, our results are more pronounced in firms that tend to attract the attention of retail investors. Our study suggests that real-name verification policies can have disciplinary effects on internet users, improving the information quality on investment-related social media. China implemented a real-name verification policy requiring internet users to register their verified real names and citizen ID numbers with social media platforms and other online services before posting content.
Research
The rise of social media platforms has created new ways for the financial markets to generate and disseminate value-relevant information to investors. Users could share their opinions anonymously, which promotes collective wisdom through facilitating crowd participation, the diversity of opinions, and freedom of bias and conflicts of interest.
However, user anonymity may also encourage extreme opinions, fake news, and rumours that can crowd the internet space and lower its information value. In response to the rising concern over misinformation on social media, some policymakers and scholars suggest removing user anonymity on social media to combat the spread of misinformation. In early 2017, China’s National Congress passed the Cybersecurity Law, which requires all website operators to verify the identity of their users using real-name registration with a citizen ID card and mobile phone number.
Can the real-name verification policy enhance the platform’s information quality? In this study, we examine the impact of removing the protection of user anonymity on the information quality of an investment-related stock platform by exploiting the recent Cybersecurity Law in China as our policy shock. This new law requires users of all platforms in the entire country, including investment-related social media platforms, to register with their real names using citizen ID cards or mobile phone numbers. Some platforms (Xueqiu, a Twitter-like stock forum) had already adopted the real-name verification policy before the policy’s effective date. This enables us to design a difference-in-differences regression by examining the adoption effects on a particular online platform (Guba, a BBS format stock forum) that enacted the verification policy in response to the new Cybersecurity Law while using another platform that had adopted a verification policy at an earlier period as a control.
Conclusion
We use the recent real-name verification policy set forth in China’s 2017 Cybersecurity Law as an exogenous shock to test whether this new policy’s effect on investment-related social media platforms is information enhancing or reducing. This setting allows us to use the difference-in-differences research design by using the posts on Guba, a popular online stock platform, after its adoption of the verification policy as our treatment sample, and the posts of the same firm during same event window from Xueqiu, another popular online stock platform that adopted the real-name verification policy in 2015, as the control sample. We find that Guba’s adoption of the verification policy has an overall positive incremental effect on its information quality, consistent with the disciplining hypothesis.
By further partitioning our Guba sample into various user groups, we find that the disciplining effect comes primarily from the information quality enhancement of continuing users who stayed on after the new policy as opposed to from naïve investors leaving the platform and being replaced by new users with better information. In particular, those continuing users post less frequently, but the information quality of their opinions becomes higher. Such disciplining effects are more pronounced for firms that tend to attract the attention of retail investors. Our results contribute to the literature examining the role of social media in disseminating value-relevant information to the financial market through the wisdom of the crowd. We show that the adoption of the real-name verification policy improves the information quality of an investment-related social media platform in China, suggesting that such policies can have a disciplinary effect on internet users. This result has implications for other developing nations with weak information environments as well as large internet platforms when considering improving their information quality.
Read the full paper.
Related Data
The social media sentiment data utilized in the study are sourced from the "Guba Analytics for China Retail Investor Sentiment" and the "Xueqiu Analytics for China Retail Investor Sentiment".
By utilizing in-house NLP models which are dedicatedly optimized for Chinese stock forum posts and trained on proprietary manually labelled and cross-checked training data, these datasets provide accurate text analytics of post content, including but not limited to quality, sentiment, and relevant stocks with relevance score. In addition to the aggregated statistics of stock sentiment and popularity, these datasets also provide rich and fine-grained information for each user/post in record level. For example, it reports the registration time, the number of followers for each user, and also the replies/readings and province being published for each post. Moreover, these metadata have been processed in a point-in-time (PIT) manner since 2019.