The evolving influence of news and social media sentiments on Indian sectoral returns during COVID-19: TVP-VAR extended joint connectedness approach

Purpose Motivated by the growing significance of news and social media-based investor sentiments in financial markets, this study examines their evolving influence on Indian sectoral returns during COVID-19. Design/methodology/approach Our study employs a robust methodological framework, including time-varying parameter vector autoregression (TVP-VAR) with extended joint connectedness and a nonparametric transfer entropy (TE). Using daily returns from eleven Indian key sectors, along with news and social media sentiment indices, we analyze both the pre-pandemic and pandemic phases from January 2018 to August 2022. Findings Our findings reveal a shift in sentiment dynamics during the pandemic, with news media changes from a net transmitter to a net receiver, while social media emerges as the dominant transmitter. Moreover, fast-moving consumer goods (FMCG), metal (MET), auto (AM), real estate (RE), consumer durables (CD) and financials (FIN) act as net transmitters, whereas pharma (PH), healthcare (HC), energy (EN) and oil and gas (OG) consistently remain net receivers across both sub-samples. Additionally, pairwise analysis reveals that social media sentiment drives EN, OG, HC, PH and IT sectors, while news sentiment remains a net receiver across all the sectors during the pandemic. Further, TE analysis shows that information flows more significantly from social media to sectoral returns than from news media during the pandemic. Practical implications Our study has important implications for sentiment-driven dynamics in the financial markets of emerging economies. Originality/value Our study examines the dynamic connectedness between media-based investor sentiments and Indian sectoral returns across pre-pandemic and pandemic periods.