Research by Adelphi University finance professor Huajing Hu, PhD, found that sentiment on X can temporarily push stock prices away from their fundamental value.
Investors make money by being patient, disciplined and well informed. But, in the end, they’re only human. Investor sentiment—whether positive or negative—can literally move markets, stirring herd behavior and sending stock prices spiraling.
Huajing Hu, PhD, professor of finance and economics at Adelphi, saw the consequences of investors’ reactions to the dot-com bubble and the 2008 financial crisis, two of the biggest shocks to the global economy in our lifetimes. “In those episodes, investor sentiment played a very important role in the valuation of stock prices,” she said.
While investor sentiment was once calculated through third-party studies, today we only have to consult social media for a real-time window into market impact. Research has shown that aggregate sentiment from social platforms can predict market-level returns more accurately than The Wall Street Journal.
Dr. Hu wondered whether social media sentiment could predict returns at the firm level—and not just same-day returns, as prior studies had examined, but over a longer period. Along with collaborators Karolina Krystyniak, PhD, associate professor at Ontario Tech University, and Hongqi Liu, PhD, of the Shenzhen Finance Institute and School of Management and Economics, The Chinese University of Hong Kong, in Shenzhen, China, Dr. Hu developed a study with the Bloomberg Twitter Sentiment Score (BTSS), which measures daily investor sentiment about individual stocks from publicly traded companies.
Every day, investors on the social platform X encounter press releases, earnings announcements and other information that shapes their feelings about their investments. The BTSS algorithm tracks investors’ responses to this news, identifying shifts in mood and tone and classifying each response as positive or negative. The BTSS then aggregates those signals into a daily sentiment score, which Dr. Hu and her colleagues averaged to a monthly score, hoping to test the algorithm’s predictive power on longer-term returns.
The team’s final dataset included nearly 4,000 firms over a period of just under three years. Their analysis showed that when investors became irrationally optimistic, stocks tended to become temporarily overvalued. This optimism was associated with higher contemporaneous returns, but those gains were later reversed as prices moved back toward their fundamental values. Conversely, pessimistic sentiment captured by the Bloomberg Twitter Sentiment Score often led to temporary undervaluation, followed by positive future returns.
A temporary misvaluation due to investor sentiment will correct itself, Dr. Hu’s team learned, but it may take some time. Once investor sentiment had artificially driven prices up or down, they looked for evidence of price reversal over the following year. To do this, they utilized what the paper calls a “novel decomposition” of social media sentiment into fundamental versus pure sentiment—in other words, breaking down investor decision-making into rational assessments versus irrational beliefs. “If a company’s earnings beat an analyst’s forecast, then investors might get too optimistic. Instead of increasing by one dollar, which is what it’s worth, the price of the stock might go up by five dollars. That’s pure sentiment.”
This lens allowed Dr. Hu’s team to “isolate the behavioral effects of investor sentiment on returns,” deepening the financial community’s understanding of behavioral forces in asset pricing. Their work was published as “What’s Trending? Stock-Level Investor Sentiment and Returns” (International Journal of Financial Studies, August 2025).
In addition, Dr. Hu found that misvaluations were most pronounced among small or young firms. These firms release less information than large, public firms, which makes each piece of news disproportionately load-bearing. “Those investors are more likely to overreact to information,” she said.
With investor sentiment increasingly disseminated across the digital realm, corporate decision-makers should treat social media discourse as a critical tool for market surveillance, Dr. Hu believes. It can also strengthen trading strategies. A shrewd investor can keep tabs on sentiment-driven pricing and buy or sell misvalued stocks at a favorable price, knowing they will eventually come back around. “When investors are reading news about their investments, it would be better to stay rational, to not be too optimistic or too pessimistic,” Dr. Hu said. “But there are some investors who will realize it’s a good time to invest.”
Frequently Asked Questions
Investor sentiment on social media can temporarily push a stock’s price away from its fundamental value. In a study led by Huajing Hu, PhD, professor of finance and economics at Adelphi University, irrational optimism was associated with temporary overvaluation that later reversed, while pessimism often led to temporary undervaluation followed by positive future returns.
The BTSS is an algorithm that tracks investors’ responses to news on X, classifies each response as positive or negative, and aggregates the signals into a daily sentiment score for individual stocks. Dr. Hu’s team averaged the daily scores into a monthly score to test their predictive power on longer-term returns.
The study analyzed nearly 4,000 firms over just under three years. Dr. Hu collaborated with Karolina Krystyniak, PhD, of Ontario Tech University, and Hongqi Liu, PhD, of the Shenzhen Finance Institute and The Chinese University of Hong Kong, Shenzhen. The findings appear in “What’s Trending? Stock-Level Investor Sentiment and Returns” (International Journal of Financial Studies, August 2025).
The study’s “novel decomposition” separates rational assessments of news (fundamental sentiment) from irrational beliefs (pure sentiment). For example, if a company beats an earnings forecast and its stock should rise by one dollar but jumps by five, the overreaction is pure sentiment.
Misvaluations were most pronounced among small or young firms. These companies release less information than large, public firms, so each piece of news carries disproportionate weight and investors are more likely to overreact.