MURFREESBORO, TN — August 2, 2017. User-generated content and attention stemming from the social media platform Twitter has wide-ranging implications for financial markets, a new study shows.
Scientists argue that Twitter activity is a source of investor attention, and is significantly associated with noise trading. As Twitter activity about a stock increases, the stock experiences a jump in returns on the following day and a subsequent reversal on day 3.
A new study explored two potential explanations for the observed Twitter attention effect: an investor recognition hypothesis and an impediments to trade hypothesis. Study results showed support for both.
The impact of Twitter attention is greatest for smaller, less visible stocks with lower analyst coverage and greater individual ownership as well as for illiquid and difficult to arbitrage stocks. This observed Twitter return impact is both statistically and economically meaningful.
Researchers generated a simple long/short trading strategy to take advantage of the “attention-premium” based on Twitter activity and stock size. This strategy resulted in a characteristic-adjusted 291 % excess return over the 5-year sample period.
In addition to an increase in noise trading, Twitter activity is also associated with permanent price impacts when it occurs in conjunction with media coverage for smaller, less visible stocks.
When a stock experiences a news release, increased Twitter activity nearly doubles the day 1 news coverage impact on stock returns with no subsequent reversal. Because the news effect is greatest for less visible stocks, Twitter assists in removing informational asymmetries often experienced by stocks that investors are less aware of.
Individual investors may have their attention drawn to specific stocks for a variety of reasons. Tips based on valuable information are only one such reason for individuals to discuss a stock. Investors may also notice stocks due to traditional new reports about the company, popular interest in a firm’s products or operations, or psychological factors triggered by some other condition.
A central issue in asset pricing has been how to distinguish between attention generated through fundamental news and the transitory interest driven by noise.
Study used social media data from Twitter to identify conditions under which retail investors’ attention on individual stocks relates to informed trading that leads to permanent price increases and when such attention triggers noise trading that results in a short-term return spike followed by a reversal.
Unlike earlier studies, Associate Professor of Finance at the University of Texas at Arlington, David Rakowski, Sara E. Shirley, Ph.D, and Jeffrey R. Stark, Ph.D., of the Middle Tennessee State University, exploited the outward-facing, rapidly-consumed, and consumer-driven nature of Twitter to explore noise trading and the spread of fundamental information in the presence of news coverage, rather than examining the overall level of trading or direction of sentiment.
Furthermore, scientists incorporated the innovations of prior research to better consider the differences between Twitter posts and other measures of attention, such as news coverage, analyst coverage, GSV, and security size. These differences reflect fundamentally different purposes, uses, content, and time frames over which attention is both disseminated and consumed.
“Prior research on Twitter and financial markets typically focuses on textual analysis of tweets to extract measures of investor sentiment, centering on text-analysis algorithms. Our approach differs, as we identify the type and impact of trading activity that occurs when attention is generated through social media activity,” say scientists.
User-generated content from Twitter impacts stock returns in ways that are distinct from other previously considered sources. Social media and online self-sharing news sources provide individuals with an important medium to generate attention and spread information about any topic, including financial markets.
Because of the large economic impact, coupled with the rising interest in social media platforms such as Twitter, it is of great interest to both academics and practitioners to better understand how individual investors can influence financial markets using nothing more than an app on their cellphone.