New Study Focuses on Social Media Data and Investor Sentiment

Finance researchers are watching with interest as postings on social media are increasingly mined for business insights. Assistant Finance Professor Samir Saadi thinks they just might be the new tea leaves of market sentiment, enabling investors to pick IPO stocks with more confidence and avoid investing in over-priced IPOs. Managers could benefit too. “Each year, IPO firms leave billions of dollars ‘on the table’ by underpricing, and pay millions in underwriting fees as a result of uncertainty around their valuations,” says Saadi, who was recently awarded a $74,718-grant from the Social Sciences and Humanities Research Council of Canada for his study, “Social media, investor sentiment, and initial public offerings.” He anticipates that as the means to exploit social media for its predictive power become more efficient and reliable, “managers will potentially save millions of dollars by reducing the risk of capital underpricing and also boosting their negotiation power vis-à-vis underwriters.”

The past five years has seen an acceleration of interest in mining social networks in both academia and industry. What began as a potentially useful vehicle to understand social interaction and information exchange is now a well-established field of research exploring the predictive power of social media and its application in a broad range of areas. One of the emerging trends in capital markets research, for example, is to investigate links between social media and aggregate stock market movement. “For the most part, though, empirical studies in finance have yet to be transformed by applying data mining to social media,” Saadi explains.

This is not for lack of interest in social media, or ignorance about its power. The IPO/capital markets sector recognizes the emerging potential to glean predictive insights from Twitter’s 645 million active users (as of January 2012), the 240 million tweets sent daily, and the more than 33 billion pieces of content shared weekly on Facebook. What the industry lacks is a reliable measure of investor sentiment. But the rise of social networks combined with advances in computer engineering has begun to open “a new frontier,” says Saadi, “in developing reliable scalable and rapid assessments of public sentiment.”

Saadi and his team propose a novel measure of market sentiment, building on recent research showing that social media can accurately capture public sentiment. The researchers will also examine the question of whether market sentiment influences post-IPO investment decisions and long-term stock performance.   

As new technologies bring new pools of social media data within reach for study, there’s a huge untapped opportunity for researchers to explore initial public offerings, as well as for firms and investors. In this last group, Samir suggests it’s the individual investors who could potentially benefit the most. In the case of IPO underperformance over the long term, they tend to be on the losing end more than institutional investors, he notes. But they might be able to select worthwhile IPO stocks with greater assurance using the new tea leaves of market sentiment.