Key points are not available for this paper at this time.
Abstract Macroeconomic factors and sentiments affect investors' decisions and thus stock returns. However, do sentiments on macro‐economic news explain stock returns? This article proposes a theoretical model to explain the relationship between stock returns and the market misperception which is driven by investors' sentiments. Then, microblogs regarding the macro‐economy posted on Sina Weibo, a mainstream Chinese Social Network Site, are extracted to measure investors' macroeconomic sentiments (IMSs) through machine learning approaches. A preliminary event study suggests that IMSs capture the development of influential macroeconomic events. Empirical results demonstrate that orthogonalized IMSs including anger, disgust, fear, joy and sadness exert heterogeneously significant effects on the Shanghai Composite Index (SHCI), and no reverse effect is found. Thus, the IMS contains additional information related to the macro‐economy; but cannot be explained by macroeconomic factors. IMSs improve the in‐ and out‐of‐sample predictabilities of SHCI returns. Thereby, investors' sentiment can be an important channel through which the macro‐economy affects the stock market.
Xu et al. (Mon,) studied this question.