This comment points out a flaw in Gau and Wang's recent empirical analysis of appraisal return bias and clarifies a point in Geltner's previous article that may have misled Gau and Wang as well as others. In examining appraisal bias in returns it is important to distinguish between cross‐sectional versus time‐series moments. In dealing with time‐series data, both the arithmetic mean and the geometric mean returns offer useful information and can complement each other in correcting for appraisal bias in the first moment of returns s