We will see evidence in Session Six that earnings surprises can be used as a statistically significant signal for post-earnings drift, such as in Rendleman, Jones, and Latane (1982). Positive earnings surprises are associated with positive alpha and negative earnings surprises with negative alpha for up to six months following the surprises. Can you suggest any reason to be cautious before accepting these results as proof that stock markets are inefficient? How would a hedge fund seek to exploit this finding, for example, and what might it cost to sustain the fund built on this play?