Review of Accounting Studies, 7, 313–318, 2002
2002 Kluwer Academic Publishers. Manufactured in The Netherlands.
Discussion of “Earnings Surprises, Growth
Expectations, and Stock Returns, or, Don’t Let
an Earnings Torpedo Sink Your Portfolio”
JOHN R. M. HAND
UNC Chapel Hill
I’m pleased to write this discussion of Skinner and Sloan (2001). It is an interesting and
well-executed study. My comments fall into three categories. First, I classify the paper in
the sense of suggesting what “type” of paper it is. This is useful because a paper’s type in
large part determines what is and what is not fair to both commend and criticize. Second, I
highlight what to me are the interesting results of the paper, and why. Finally, I raise three
small criticisms of the paper.
1. Types of Papers
Oversimplifying, in my view there are ﬁve major types of papers in accounting research.
T1 Papers that ask “Is XYZ true?”
T2 Papers that take XYZ to be true, and ask “Why is XYZ true?”
T3 Papers that take XYZ to be true, and ask “Given XYZ, what does that imply for ABC?”
T4 Papers that ask “What happens if . . . ?”
T5 Papers that distill reality into analytically tractable models.
The paper by Skinner and Sloan is a combination of paper types T1 and T3. In terms of T1,
Skinner and Sloan ask the question “Do investors in growth stocks systematically assume
that the earnings of such stocks will be higher than they actually are?” They conclude that the
answer is yes, showing that growth stocks are more likely to experience negative earnings
surprises and that for growth stocks the reaction to negative earnings news is more negative
than is the positive reaction to equivalent amounts of positive earnings news (Figure 4).
Insofar as T3 is concerned, Skinner and Sloan ask whether the systematically negative price
revisions that occur when growth stocks’ actual earnings are made known fully explain the
low returns of growth stocks in comparison to value stocks. Here too they conclude that
the answer is yes, demonstrating quite convincingly that after controlling for the asymmetric
price response of growth stocks to negative earnings news, there is no evidence of a return
differential between growth and value stocks over the long-run.