Sentiment on the stock markets: Evidence from the wavelet coherence analysis

Abstract

In this paper we propose to use wavelet analysis to determine the relationship between investor sentiment and stock index. Based on stock indices in various markets and alternative indicators, sentiment using wave analysis uses the intensity and correlation between sentiment and stock index returns in the short and long term. Initially a strong relationship identified by wavelet coherence and phase difference gradually becomes less intense with increasing time horizon. This suggests that overestimation or underestimation in the short term is gradually corrected in the long run. In addition, a stronger relationship can be observed between investor sentiment and stock indices in times of crisis, including COVID-19 pandemics.