When Investors Can Talk to Firms, Is It a Meaningful Conversation? Evidence from Investor Postings on Interactive Platforms

Abstract

We investigate postings on two unique online interactive platforms by investors around earnings announcements of publicly listed firms in China. We find posting volumes on the platforms increase around earnings announcement dates, suggesting that investors acquire firm-specific information via the interactive platforms. This relation is stronger for firms with higher reply rates or more timely replies, for firms with fewer shares held by institutional investors, for firms with larger information asymmetry, and when market uncertainty is higher. Furthermore, the positive association between posting volumes and earnings announcements is more pronounced for firms with smaller Baidu search volumes. This suggests a substitution relationship in collecting information between asking listed firms questions directly and searching online. Finally, our evidence suggests that questions posted around earnings announcements accelerate the price discovery of earnings and attenuate post-earnings announcement drift and stock price synchronicity.

Keywords:

• Information acquisition
• Internet
• Interactive platform
• Posting
• China

JEL classifications:

• M41
• G14
• D83
• G30

Acknowledgements

We thank Jeffrey Ng (the editor), two anonymous referees, Steven Cahan, Nafiz Fahad, Barry Oliver, Phillip Stocken, Eric Tan, Mark Wilson, Bohui Zhang, and participants at the 2021 Accounting and Finance Association of Australia and New Zealand (AFAANZ) Conference and seminar participants at The University of Queensland and Central South University for their valuable comments.

Data Availability

The data analysed in this study are derived from the China Stock Market & Accounting Research (CSMAR) database and the following resources available in the public domain: Hu Dong Yi: http://irm.cninfo.com.cn/, e Hu Dong: https://sns.sseinfo.com/, and Baidu index: http://index.baidu.com/v2/index.html#/.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplemental Data and Research Materials

Supplemental data for this article can be accessed on the Taylor & Francis website, https://doi.org/10.1080/09638180.2022.2118147.

• A1. Additional Analyses: Posting around Various Events

• A2. Additional Analyses: Comment-Based Posts

• A3. Robustness Tests: An Exogenous Shock

• A4. Robustness Tests: Various Time Windows During the Event

Notes

1 For example, Jung et al. (2018) find that firms are less likely to disseminate information via Twitter when the news is bad.

2 Hu Dong Yi: http://irm.cninfo.com.cn/.

3 e Hu Dong: https://sns.sseinfo.com/.

4 The official regulations are at http://www.tzyhj.cn/rule/11/gyszzqjysssgstzzgxhdpthdyygsxdtz.html and http://www.sse.com.cn/lawandrules/sserules/listing/stock/c/c_20150912_3985864.shtml.

5 The digital identification systems on the platforms automatically exclude posts regarding advertising, repeated questions from the same person, incomplete postings, and any postings that violate regulations.

6 Examples of those who break the rules of the platforms and are punished by the Shenzhen and Shanghai Stock Exchanges are listed here: https://www.sohu.com/a/298300197_733484.

7 A-share firms listed on the Shanghai Stock Exchange start in 2013 because e Hu Dong was launched on 5 July 2013.

8 We manually read through all the posts to classify whether a post is a question or a comment.

9 The website is http://index.baidu.com/v2/index.html#/.

10 The small mean value of the number of daily posts indicates that investors do not frequently ask questions via the online platforms. One possible explanation for this observation is that all historical questions and replies are available to the public. Moreover, for investors’ convenience, each platform provides a key-word search function to locate specific questions or replies. Investors can go through the historical information to find answers, rather than posting a duplicate question. In addition, for repeated questions, firms might reasonably ignore them, copy previous answers, or tell investors that this question has already been posed previously and refer them to the prior reply, which has the effect of limiting the posting of repeated questions.

11 One issue is that the results seemingly contradict each other; this will be a real concern if our sample has a large overlap between firms that are more responsive on the platform (i.e., disclose more information) with firms that have a higher financial reporting opacity (i.e., hide information). We thank an anonymous referee for raising this issue. To address this concern, we partition our sample observations into two groups based on annual median value of financial reporting opacity and find that firms with higher financial reporting opacity are associated with a lower reply rate and longer reply time, thereby, consistently showing that firms with higher financial reporting opacity hide information from outside investors. Therefore, there should not be a large overlap between the two subsamples. Nevertheless, given that an overlap could bias our results, we further add ‘reply rate’ and ‘reply time’ as additional controls in regression model (1). In addition, we re-estimate regression model (1) for two subsamples of observations—namely, high versus low financial reporting opacity. The results (untabulated to conserve space) which remain the same, qualitatively; are available from the authors upon request.

12 The assumption underlying these analyses is that investors can obtain firm-specific information via posting questions on the platform. This assumption is reasonable given that there are several policies to facilitate listed firms’ active participation in the online platforms as discussed in Section 2.1. In addition, our descriptive statistics show a high reply rate in our sample—a reply rate exceeding 85%. Furthermore, we find a significant positive relation between investors’ posts and the replies from listed firms, further validating this assumption.

13 We also utilize the three-factor model proposed by Fama and French (1993) to calculate the predicted normal daily return. The results remain qualitatively similar to those reported in Table 6.

14 Regarding firms in the manufacturing industry (C), we use the two-digit CSRC industry code (C1–C9).

Additional information

Funding

Shijun Guo acknowledges financial support from the Fundamental Research Funds for the Central Universities (grant number 2022CDSKXYJG007) and China Postdoctoral Science Foundation (grant number 2022M710511).