An important challenge for investors is understanding why some stocks and investments perform better than others. Traditionally, financial experts have predicted how much return (or profit) an investor can expect based on a company’s size, its earnings, or overall economic conditions. With important implications for investors and other finance professionals, Chu Zhang of HKUST introduces a new way of looking at pricing errors—focusing on the differences between expected returns and actual returns.

Examining these errors, Zhang shows that traditional models used to predict investment performance might not be as accurate as previously thought. For example, the author notes, “macroeconomic variables, though intuitively appealing, do not work well in practice because they are not adequately reported in time to be aligned with return horizons except for interest rates.” Meanwhile, “the statistical approach captures many variations in the returns by design,” Zhang points out, “but the extracted factors typically lack economic interpretations, and they do not necessarily explain expected returns.”

In short, many existing models overlook important details, particularly how different firm characteristics or traits can affect a stock’s risk and return. Helping to address these limitations, Zhang proposes a new model in which “pricing errors (alphas) are specified to be orthogonal to the affine-transformed firm characteristics.” This specification diverges sharply from the literature, “as the zero pricing error hypothesis is strongly rejected for various models with commonly used firm characteristics.” Through this innovative framework, the author redresses the weaknesses of previous predictive models and sheds light on the consistency (or otherwise) between firm characteristics and expected returns.

This work paves the way for better informed financial decisions, offering valuable tools for practitioners and academics alike in navigating the complexities of asset pricing.