Pssst… we can write an original essay just for you.
Any subject. Any type of essay.
We’ll even meet a 3-hour deadline.Get your price
121 writers online
Financial behavior analysis indicates a relationship between investors sentiments and stock prices. In this study, I examine Tweets, by sentiment, specifically tagged with a company’s financial ticker. I analyze roughly 336,317 microblogging daily messages to capture investors’ sentiments related to S&P100 companies for predicting varied price movements. I find evidence that bullishness of stock microblogs is positively and significantly correlated with abnormal return. Consistent with previous studies, the results show that Twitter messages convey private information. Evidence shows that microblogging volume can predict trading volume during daily trading hours. The results suggest microblogging features can be valuable real-time proxies for investor decision-making. There is evidence of a positive and significant correlation between bullish tweets and CARs for firms experiencing an income increase compared to prior year. Firms that experience an income decrease show a positive and significant correlation between bullish tweets and CARs during normal trading hours, and a negative and significant correlation between the bearish tweets and CARs during normal trading hours. All these results imply that people use Twitter to obtain information and financial interpretations from peers and experts in times of economic uncertainty.
The analysis of stock market behavior has been the focus of numerous researches in academia. The study of behavioral finance has been growing and is considered one of the most promising areas to understand investors and participants behavior and the impact of this behavior on financial markets. A model that can analyze the investors’ sentiments and attention could provide significant predictive power when attempting to forecast stock market volatility and perhaps even when modeling stock market return. Qiu and Whinston (2011) find that short-term sentiment plays a vital role in the short-term performance of stocks in financial markets. Additionally, studies have shown that investors emotions have significantly affected their financial decisions. These emotions can be derived from social media activity. Investors use social media platforms to discuss public companies and trading ideas. The attraction of social media research has increased in both academia and industry due to the vast amount of inferential data. Since Twitter was founded in 2006, it has become the most important communication tool between organizations and investors. There are communities within Twitter that share information relative to their interests, one of these communities has developed around the stock market where investors and traders share information and trading ideas about the stock market. Using Twitter sentiments to predict a stock or sector movements may yield into potential applications that provide a decision support mechanism for traders and investors. Bollen, Mao, and Zeng (2010) analyzed the Twitter population “mood” to predict the changes in Dow Jones Industrial Average (DJIA) closing values with a reported 87.6% daily price predicting accuracy. They utilized an immense amount of data in order to understand the users’ overall sentiments as a proxy for investors’ sentiment about the stock market. Vu, Chang, Ha, and Collier (2012) studied consumer satisfaction sentiments and stock price movement for four large Tech companies’ products. They documented accuracy that ranges between 75% to 82.93% in predicting the daily up and down changes of Google, Apple, Microsoft, and Amazon. In this paper, I will examine investors sentiments on Twitter for abnormal stock return rather than consumer satisfaction.
H1: Bullish (Bearish) tweets in day t are positively (negatively) correlated with abnormal return in day t+1.
Wysocki (1998) documented that overnight message-posting volume can be used as a proxy to predict next-day stock trading volume and price movements. Also, he found that the cumulative posting volume is high for firms with high volatility and trading volume. Antweiler and Frank (2004) documented a positive relationship between activity in Internet message boards and stock volatility and trading volume. Saavedra, Hagerty, and Uzzi (2011) “examine the association of synchronous trading with individual performance and communication patterns”. They find that the higher synchronous trading the better performance realized at the end of the trading day. They documented a positive relationship between instant messaging patterns of traders and the level of synchronous trading. Brown (2010) documented a positive relationship between the volume of tweets posted during the stock market overnight closing period and the next day trading volume. This study examines whether tweets posted on discussion boards during trading hours have any effect on the trading volume before the market is closed (between 9:30 a.m. and 4:00 p.m. weekdays).
H2: Twitter message volume is positively associated with stock trading volume during the market operating hours within the same trading day.
Financial market movements are driven by participants beliefs and actions towards future outcomes, and aggregate societal sentiments should be a reliable predictor. However, not all social media users are equally influential. Users with greater influence on social media have a greater impact on their community sentiments. Some users on Twitter are more informed with expert opinions on the stock market than others. Influential users and experts could themselves affect the prices by inducing other investors trading behavior or beliefs, therefore, identifying influential users is useful to predict asset price movements. Investment advice provided by some users with a high number of followers and/or a high number of retweets (quotes) will have more influence than other users.
H3: The sentiments of influential users “measured as the number of retweets” have an impact on the stock price.
Wysocki (1998) tests whether the variation in message-posting volume is related to numerous firms’ performance attributes and find that the cumulative posting volume is highest for firms with high extreme past return and accounting performance. Therefore, I expect that Twitter message volume will increase around the earnings’ announcement, and the overall sentiments will impact the abnormal return simultaneously with the earnings surprise relative to the past performance.
H4: Stock return following firms’ earnings announcements are positively correlated with investors’ sentiment on Twitter.
The purpose of the current study is as follow. First, I replicate previous studies and examine whether Twitter will influence the next-day stock return while controlling for the firm’s risk factors. This study considers the firm’s characteristics in analyzing the corporate microblogs to predict stock price movement. Considering the firm characteristics and its financial announcements give more accurate and reliable results. This issue is of key importance since other studies, rely only on analyzing Twitter influence in the price movements regardless of the firm’s performance and its financial announcements which fundamentally determine an important portion of the stock return. Second, I study investors sentiments in the return-generating process and investigate whether the market fully realizes investors sentiments simultaneously during the trading hours within the same trading day. This issue is important to examine the timeline of Twitter influence on investors trading behavior. Third, I investigate the role of investors sentiments on stock price movements around firms’ earnings announcements.
To my best knowledge, this study is the first one to examine Twitter messages influence on the stock return around firms’ financial (earnings) announcements. There is evidence that bullishness of stock microblogs is positive and significantly correlated with abnormal return. Twitter stock blogging conveys private information, this is consistent with a previous study. There is evidence that stock microblogging volume is correlated with the trading volume during the trading hours. This evidence suggests microblogging features can be valuable real-time proxies for investor decision-making. Results show a positive and significant correlation between bullish sentiments and cumulative abnormal return (CARs) around earnings announcements. Also, a positive and significant relationship between bullish sentiments and CARs during the trading hours for firms that experience an increase or decrease in their income relative to past performance. For firms with decreasing income, I find a negative and significant relationship between bearish sentiments and CARs during the trading hours. Literature review Researchers buy into the idea that investors look to the opinions and other information from peers through social media to fuel their investment decisions; Chen, Liu, and Zhang (2012)). Peer advice influence investors who are less informed on social media.
In the stock market, a wave of investors sentiments could travel between investors resulting in significant changes to a stock price. These waves of communication are signs of new information flowing into the market. The share and trade of information and investing ideas among investors effects the stock price movement and trade volume. Sprenger and Welpe (2010) find a relationship between tweets volume and the trading volume. Oliveira, Cortez, and Areal (2013) measured the posting volume associated with tweets related to firms and find that sentiment indicators can explain stock return. Zhang, Fuehres, and Gloor (2011) find a negative and significant correlation between emotional tweet percentages and Dow Jones, NASDAQ, and S&P 500, but a positive and significant correlation to VIX.
Efficient Market Hypothesis states that stock price movements are driven by new information and these movements will follow a random walk pattern rather than following past and present prices. However, Nguyen, Shirai, and Velcin (2015) argue that historical prices with information from social media and sentiment analysis can improve stock price prediction models. They build a model that predicts stock price direction with a claim of 54.41% accuracy. Sample Formation Twitter API is used to collect data. This data includes many features of Twitter including individual Tweet content, timestamps, number of retweets, and the users’ information (Brown (2010); Twitter (2011a)). All tweets considered used the English language and are posted during the period between January 1st, 2013 through June 30th, 2013, excluding the weekend days and holidays were stock markets are closed. Additionally, I used only Tweets mentioning the “$” nomenclature which is widely used in Twitter investing and trading communities for discussions related to a specific company’s share. For example, to track Amazon stock on Twitter, a “$” would be added to the Amazon stock symbol, $AMZN instead of simply AMZN. Using this method will reduce the number of irrelevant messages, resulting in a noise-minimized source of data in the analysis. This also acts as a proxy to limit the analysis to more influential users. The original Tweet time has been normalized to the American Eastern Standard Time.
This study focuses on S&P100 which include well-known companies that attract a large number of mentions in social media. However, forty-one firms of S&P100 were excluded from the sample due to limited tweet availability during the study period. Table IV shows a summary statistic of each firm’s number of total tweets, total positive, negative and neutral tweets, return and trading volume. The sentiment categories that were used are bullish for tweets that refer to positive sentiments, bearish for tweets that refer to negative sentiments, and neutral for the tweets that do not convey any sentiments . Finally, the stock market daily returns data was collected from the Center for Research in Security Prices (CRSP) files, and the firm’s characteristics data were collected from Compustat files.
Investors Sentiments and Abnormal Return Stock price movement prediction has been a source of interest for many academic researchers. Although many researchers have attempted to scientifically predict stock price movement, no method, up to this date, has proved successful. The complexities associated with different parameters that are constantly shifting make stock price movement difficult to predict. The efficient market hypothesis (EMH) suggest that the financial markets are “informationally efficient” meaning that prices reflect all known information. Corporate insiders have private information that leads to excess profit, but outsiders cannot earn profit from trading strategies based on public information. One current popular method of stock market volume and return prediction comes from investors sentiments analysis. “Short-term sentiments play a very important role in short-term performance in financial market instruments such as indexes, stocks and bonds”.
Investors’ emotions can affect the way they react to new information and can impact their decisions which explains why their decisions sometimes depart from the rational behavior. Microblogs are an ideal early warning of stock price movement as they provide a simultaneous update and spontaneous glimpses into investors’ opinions and sentiments about their future trading behavior.
To export a reference to this article please select a referencing style below:
Sorry, copying is not allowed on our website. If you’d like this or any other sample, we’ll happily email it to you.
Your essay sample has been sent.
Want us to write one just for you? We can custom edit this essay into an original, 100% plagiarism free essay.Order now
Are you interested in getting a customized paper?Check it out!