Price momentum is frequently used to predict future returns. But there is no consensus theory on why it works. In this study, we show the limitations of using price momentum as a signal of future returns.
We find that three measures of information momentum: analyst estimate revisions, tone changes in earnings conference calls, and tone changes in Bloomberg News sufficiently capture the predictive ability of price momentum. This evidence leads us to advocate for the use of information momentum in investment decisions rather than price momentum alone.