Memo
DATE: January 19, 2015
SUBJECT: How Do Earnings Numbers Relate to Stock Returns?
As for the motivation, in the past literature, it developed a relation between accounting earnings and stock returns. This article follows the steps of former articles and provides deeper insights for understanding this relation. It developed a large body of theory about the association between accounting income numbers and changes in share prices. But there are still some questions that haven’t been solved. In this article, the authors tried to figure out what kinds of questions are still unexposed and what new research methods they developed.

As for the description of the data used, researches uses “Three Links”. Researchers announced that there are factors that can impact Three Links .Then, researchers used samples of firms listed on the NYSE, AMEX, and NASDAQ exchanges. Researchers required their data from the S&P Compustat database and the CRSP database. For persistence and association test, researchers used annual earnings and returns data for 31,923 firm-year observations over 1988 – 2001. For efficiency test, they contained quarterly earnings and daily returns data for 90,470 firm-quarter observations. Researchers’ analyses use accounting earnings information to explain differences in firms’ stock returns. Researchers’ final analysis examined two dimensions of market efficiency with respect to earnings information. They again examined daily stock returns surrounding earnings announcements, collected daily returns from CRSP and computed cumulative abnormal returns through each day. They made similar computations of cumulative abnormal returns. They grouped firms into deciles based on unexpected earnings per share and summarized each decile portfolio’s performance by averaging the abnormal returns for the firms in each decile each quarter. They then average each decile-quarter’s return across the 58 quarters in their sample.

The paper provides three different empirical evidences on the association between earnings and returns; the effects of earnings persistence on the earnings-returns association; and the quickness and completeness of the market’s reaction to earnings news.
The first set of empirical test replicated Ball and Brown (1968), showing a strong relation that difference in the sign of the change in earnings relates to a difference of over 35% in the average firm’s annual stock returns. Also, the paper extends the relation between the magnitude of the change in earnings, operating cash flow versus the stock returns.
The second set of empirical test replicated Kormendi and Lipe (1987) by using the new data of the sample firms. The paper supports that stock returns associated with earnings increases are much higher for firms with high earnings persistence than for firms with low earnings persistence. Beside, the paper predicts the relation between earnings decrease and firm’s earning persistency.
The third set of empirical test replicated Bernard and Thomas (1989). First, the results show that stock returns has strong relation with quarterly earnings information, even anticipate earnings information. Second, the results show that stock returns are