Individual Investor Sentiment and Stock Returns - What Do We Learn from Warrant Traders?

Abstract

In this paper, we propose a measure of individual investor sentiment that is derived from the market for bank-issued warrants. Due to a unique warrant transaction data set from a large discount broker we are able to calculate a daily sentiment measure and test whether individual investor sentiment is related to daily stock returns by using vector autoregressive models and Granger causality tests. Our sentiment measure should be, ex ante, the preferred measure when compared to other sentiment measures proposed in the literature that are based on the closed-end fund discount, stock market transactions, the put-call ratio or investor surveys as it circumvents problems that are associated with the measures used in prior studies. We find that there exists a mutual influence between sentiment and stock market returns, but only in the very short-run (one and two trading days). Returns have a negative influence on sentiment, while the influence of sentiment on returns is positive for the next trading day. The influence of stock market returns on sentiment is stronger than vice versa. Our results cast doubt on whether sentiment measures are useful to predict the market over horizons of more than one day. Even our improved measure is hardly able to do so. Nevertheless, sentiment measures provide evidence on how investors trade and which factors influence their expectations.