Connectedness of Agricultural Commodities Futures Returns: Do News Media Sentiments Matter?

Using the novel daily commodity-specific Thomson Reuters Market Psych sentiment data derived from news, social media, press releases, and regulatory filings, this study investigates the asymmetric impact of news and social media sentiment on the futures return connectedness of agricultural commodities. We construct time-varying connectedness measures for agricultural commodities futures returns at different quantiles and show how these spillover measures depend on news sentiment under extreme events. Our results show that the impact of news media sentiment on agricultural commodity connectedness depends on the quantiles (lower, median, upper) and the type of sentiment (traditional news or social media). In particular, we find that social media sentiment has a statistically significant impact on the magnitude of shocks each commodity transmits to others, at both the lower and upper quantiles, indicating that the media sentiment effect is more substantial during extreme market periods.