Temperature and trading behaviours

Abstract

It is well known that the weather has an impact on human behaviours. Motivated by the extant literature concerning the positive linear relationship between temperature and investors' trading based on regional data (e.g., Schmittmann et al., 2015), we re-examine the temperature effect by utilizing a dataset that encompasses a large number of individual investors' accounts and a wider range of weather conditions, and construct a comprehensive measure of sensed temperature, Apparent Temperature, which incorporates atmospheric temperature, relative humidity and wind velocity. Our results demonstrate that the relationship between Apparent Temperature and individual investors' purchasing tendency displays an inverted U-shape. At a comfortable temperature, investors are more aggressive and tend to buy more stocks relative to selling. However, the relationship between Apparent Temperature and trading volumes displays a U-shape, suggesting that investors trade less on days when the temperature is comfortable, reducing non-financial opportunity costs. Further, Apparent Temperature displays greater external validity than the three weather conditions when they are examined individually. Our study provides original evidence of non-linearity, instead of the linearity documented by previous research, in the relationships between temperature and retail investors' trading behaviours. Our findings contribute to the literature regarding the weather-induced sentiment misattribution of a diversity of investors.