DOES TWITTER HAPPINESS PREDICT BULLISH MARKETS? A STUDY OF THE US STOCK MARKETS

This study investigates whether Twitter happiness, measured by the Twitter Sentiment Index (TSI), predicts stock market performance across seven US indices. Unlike previous studies that primarily assess linear relationships, it uses a dataset of seven US stock market indices and applies cointegration techniques and a nonlinear autoregressive distributed lag model to capture both short-run and long-run effects of Twitter sentiment on stock markets. The findings suggest that Twitter happiness has a substantial short-term impact on stock market indices while exerting a moderating effect in the long term. Notably, the influence of positive Twitter sentiment on US stock market performance is more pronounced than negative sentiment, emphasising the asymmetric nature of this relationship. Two key findings emerge: (i) a dynamic nonlinear cointegration exists between the TSI and stock market indices, and (ii) Twitter happiness has a pronounced short-term impact with a moderating effect in the long run. These findings have implications for investors, policymakers, and researchers. The TSI, derived from the Hedonometer, provides a unique and accessible proxy for investor sentiment by measuring the happiness quotient of Twitter users, offering an advantage over traditional sentiment indices. We have shown how the investors’ mood using Twitter feeds provides an automatic, fast, and accessible tool, which may even be optimised, to study the impact on stock market behaviour. One limitation of this study is its sole focus  on the stock markets in the United States, where Twitter serves as a significant news and social media platform. Additionally, the user base on Twitter consists of English speakers, predominantly in the US. To enhance the scope of this research, it could be beneficial to explore its applicability in emerging markets where Twitter is gaining traction as a social media tool.

IMR-IKS Call for Papers