Investor Sentiment and Stock Price Premium Validation with Siamese Twins from China

We apply the principal component analysis (PCA) and the partial least squares (PLS) methods to construct investor sentiment measurements and decompose them into total, local and global indices for the Mainland China (A-shares) and Hong Kong (H-shares) stock markets. Our results show that total investor sentiment is a strong predictor for the A-shares market, but it has hardly any effect in the H-shares market. We find that the sentiment differential, either total or local, has a positive relationship with the aggregate excess returns differential between the two markets, but the Hong Kong sentiment alone, either total or local, does not play much role. We further find that the sentiment differential is positively associated with the price premium of A-shares over H-shares. Our results are robust even after we add traditional factors.