HDR Candidate: Zin Yau Heng

Milestone: Mid-candidature Review

Title: Effects of information on the stock markets through the online social media network

When: Thursday, 25 March 2021; 11am -1pm.

Where: Zoom


This thesis uses a proprietary StockTwits dataset to investigates the information transmission in the online financial social media network platform. This chapter uses StockTwits dataset to investigate the impacts of activities from StockTwits on stock returns. However, A growing concern about the manipulative scheme involved from social media. First, this paper questions whether activities of social media can affect stock prices? The event study shows an early spike in the abnormal return of stock prices. The early spike indicates a market inefficient and possible stock price manipulation. The market inefficient are interrelated to stock price manipulation. Given the possible price manipulative scheme, this paper further evaluate whether the sizes of the stocks are vulnerable to the activities of social media? The findings from both events study and contemporaneous regression show that the small stocks are more sensitive to the activities of StockTwits. Therefore, the senders possibly send strategic tweets on the small stocks for personal manipulative incentives. Inspired from the first essay, the small stock is sensitive to the activity of StockTwits.

The second essay attempts to examine whether the sender sends more bullish sentiment on the small stock and whether there are heterogeneous senders send different information sets through StockTwits. We use the DID regression to shows that tweeters send a more bullish message on the small stock. The result also shows that there are 2 possible types of senders: Strategic and Neutral. The Neutral senders always send neutral messages. the result show short term investor as potential neutral senders. On the other hands, the result shows senders with intermediate as potential strategic senders. The finding attempt to inform the investor and policymaker that the information is full of noisy. The strategic senders do not always send the biased message. The strategic senders may occasionally send neutral messages to accumulate reputation. Besides, neutral senders may always send neutral messages to accumulate a reputation.

Since early January 2021, a group of users on Wallstreetbets, a subreddit of Reddit1had been pushing up the stock price of GameStop, despite its weak fundamentals. This coordinated attack against short sellers of GameStop eventually drives GameStop’s stock price up by 1500%, causing the bankruptcy of some hedge funds and legislative concerns (Li,2020). The third essay raises a How does this GameStop Works? The coordination game can explain well this scenario. The idea of understanding speculative attacks as a coordination game can be traced back to Morris and Shin (1998), who apply the equilibrium selection technique from the global game literature (Carlsson and Van Damme, 1993) to currency attacks. They show that there exists a unique cut-off in the fundamental value of the currency such that an attack will be successfully coordinated whenever the currency fundamental value falls below this cut-off. Their results imply that there may be no attack even though the currency is “ripe for the attack” (i.e., the attack will be successful as long as a sufficient proportion of investors attacks). From potential attackers’ point of view, then, there is uncaptured gain and hence incentives to improve.