A Study of Investor Sentiment in the Indian Market.

Abstract

Market sentiment has always been given importance by investors in their decisions. Classical finance theories based on investor rationality, however, largely ignored the possible role of market sentiment in asset pricing. The recently developed finance theories based on investor irrationality attempt to formally define and model sentiment and examine its impact on asset pricing. Investor sentiment is a measure of the influence of irrational behaviours on asset prices at a point of time. The sentiment prevails as rational investors fail to dominate the irrationals through arbitrage. Sentiment is often measured based on inputs from consumer confidence indices and surveys of investor mood. Financial market data such as advances to declines, dividend premium, turnover etc. have also been used to develop sentiment indices. We attempt to develop a sentiment index for the Indian market at a monthly frequency based on financial market data. The financial data used include premium on dividend paying stocks, stock market turnover, market volatility, net investments by foreign institutional investors and the number of IPO's. The sentiment index, developed from the first principal component of these variables explains 46.8% of the combined variance. We employ the sentiment index to empirically examine a number of the features of the asset prices in India. We test the hypothesis that hard to value and difficult-to-arbitrage stocks are more influenced by irrational sentiment. We expect highly volatile, small, relatively young, non-dividend paying and unprofitable stocks (characteristics of hard to value and difficult to arbitrage stocks) to be prone to sentiment. Our results show that larger, less profitable and less dividend paying stocks are more sensitive to sentiment in India. We also examine whether incorporating sentiment in asset pricing models better explains financial anomalies. Using a conditional asset pricing model and a two-step regression methodology, we find that inclusion of sentiment is able to better explain asset pricing anomalies.