Is evidence of portfolio reversals evidence of stock reversals?
For some reason, many academic papers document portfolio reversals and then claim such evidence implies something about stock reversal. The former is about the time-series, the latter is about the cross-section. For example, this is the case in the literature on S&P-500 changes, where the theory claims to be about individual stocks, but the evidence brought to bear about this is not.
Ergo: This empirical evidence, which appears in very prominent papers, is simply uninformative and wrong. This is not to say that the empirical evidence contradicts the hypotheses. It just doesn’t say much about it.