Modeling a bi-directional sentiment-return relationship: Evidence from the Indian market

Abstract

In the last two decades, the subject of investor sentiment has attracted the attention of researchers across the globe. This study attempts to examine the bi-directional relationship between investor sentiment and stock market returns in the Indian market by focusing on both contemporaneous and lagged relationships between investor sentiment and market returns. It also attempts to study the effect of lagged market returns on the current market returns. This study constructs an investor sentiment index for the Indian market using the principal component analysis technique. The results of the regression analysis between the investor sentiment index and stock market returns establish that current sentiment positively affects current market returns, and one-month lagged sentiment negatively affects current market returns. Further, it is found that a one-month lagged market return has a positive association with the current market returns. Moreover, using the VAR model, this study found the existence of a contemporaneous and lagged bidirectional relationship between investor sentiment and market returns. The results of impulse response analysis and variance decomposition analysis also support the presence of a sentiment-return bidirectional relationship but show that the effect of sentiment on market returns is more pronounced than the effect of market returns on investor sentiment.

Keywords

• Geweke test
• Granger causality
• impulse response analysis
• principal component analysis
• variance decomposition analysis
• vector autoregression