Investor Sentiment and (Anti) Herding in the Currency Market: Evidence from Twitter Feed Data

This paper establishes a direct link between (anti) herding behavior in currency markets and investor sentiment, proxied by social media based happiness index built on Twitter feed data. Our analysis of daily data for nine developed market currencies suggests that the foreign exchange is generally characterized strong anti-herding behavior. Utilizing quantile-on-quantile (QQ) approach, Sim Zhou (2015 Sim, N., H. Zhou. 2015. “Oil Prices, US Stock Return, Dependence Their Quantiles.” Journal Banking & Finance 55:1–8. doi:10.1016/j.jbankfin.2015.01.013[Crossref], [Web Science ®] , [Google Scholar]), we show relationship sentiment fact regime specific, with particularly prominent during states extreme sentiment. The effect stronger bullish states, while average associated less severe anti-herding. findings lend support to behavioral factors asset pricing models suggest real time signals can be utilized monitor potential speculative activities market.