Firm-specific investor sentiment and daily stock returns in Finland

Abstract

Classical finance literature assumes that most investors are rational utility maximizers and that arbitrageurs trade away the possible mispricing in stock markets. Thus, stock prices reflect all available information and accurately estimate their fundamental value at any given time. In contrast, behavioral finance assumes that investors are limitedly rational in their decision-making due to cognitive and emotional weaknesses. Also, arbitrages have some limitations, which cause stock prices to contain inefficient factors. Investor sentiment can be seen as one of these factors. This thesis aims to examine if investor sentiment affects stock market returns in the Finnish stock market. The relationship between these two is examined by constructing a firm-specific sentiment index using four sentiment indicators. The indicators are the relative strength index, psycho-logical line index, adjusted turnover rate, and the logarithm of the trading volume. These indicators are regressed against the market and firm variables to capture the irrational effect of sentiment. Previous research suggests that investor sentiment and stock returns have a positive relationship and that the relationship is more sensitive to firms that are harder to valuate. This thesis examines this relationship by constructing portfolios based on variables that have been found to be essential in firm valuation. Additionally, the effect of high sentiment periods is examined due to the previous evidence of a more noticeable sentiment effect on these periods. The findings of this thesis indicate that investor sentiment and stock returns have a significant and positive relationship, which implies that investor sentiment has predicting power of short-term stock returns. Moreover, the relationship between these two is much stronger for stocks that are seen as more challenging to valuate, for example, stocks of those companies that are smaller, more volatile, and with growth potential. Additionally, findings suggest that the sentiment effect is more noticeable during periods of high sentiment.