Custom Search

Price and Earnings Momentum: Multivariate Analysis

The evidence in the last section indicates that each of the momentum strategies that we consider is by itself useful in predicting future stock returns. We now examine whether the continuation in past price movements and the underreaction to earnings news are the same phenomenon.

A. Two-way Analysis of Price and Earnings Momentum

Our first set of tests addresses this issue in terms of a two-way classification. At the beginning of each month, we sort the securities in the sample on the basis of their past six-month returns and assign them to one of three equally-sized portfolios. Independently, we sort stocks and group them into three equally-sized portfolios on the basis of the most recent earnings surprise. Under this procedure each stock is assigned to one of nine portfolios. Table VI reports buy-and-hold returns over each of several periods following portfolio formation, as well as the average earnings surprise over the first subsequent year. Panel A reports the results when earnings surprises are measured as abnormal returns around earnings announcements, while Panels Р’ and РЎ provide results for standardized unexpected earnings and analyst revisions, respectively.

The first three panels in Table VI tell a consistent story. Most important, past realizations of six-month returns and earnings news predict continued drifts in returns in the subsequent period. In particular, the two-way sort generates large differences in returns between stocks that are jointly ranked highest and stocks jointly ranked lowest. For example, using past return in conjunction with earnings surprise measured as the abnormal announcement return, the highest-ranking portfolio outperforms the lowest-ranked portfolio by 7.9 percent in the first six months. Similarly, the six-month spread is 8.1 percent using prior return together with SUE, and 8.8 percent using prior return with analyst revisions.

Each variable (prior return or earnings surprise) contributes some increВ­mental predictive power for future returns, given the other variable. In Panel A, holding prior return fixed, stocks with high past announcement return earn in the first six months following portfolio formation 2.8 percent more on average than stocks with low past announcement return.9 In comparison, the returns on stocks with high and low past prior return, but similar levels of announcement return, differ on average by 4.6 percent.

Using measures of longer-term earnings news, as given by either standardized unexpected earnВ­ings or revisions in consensus estimates, turns out to place earnings momenВ­tum on a more equal footing with price momentum. The six-month spreads induced by past SUE or past revision, conditional on prior return, are 4.3 percent and 3.8 percent, respectively. Sorting on past return, conditional on past earnings news, produces average spreads in six-month returns of 3.1 percent (Panel B) and 4.5 percent (Panel C). The bottom line is that although the ranking by prior return generally gives rise to larger differences in future returns, neither momentum strategy subsumes the other. Instead, they each exploit underreaction to different pieces of information.

Table V

Mean Returns and Characteristics for Portfolios Classified by

Revision in Analyst Forecasts

At the beginning of every month from January 1977 to January 1993, all stocks are ranked by their moving average of the last six months' revisions in mean I/B/E/S estimates of current fiscal-year earnings per share, relative to begin ning-of-month stock price, and assigned to one often portfoВ­lios. The assignment uses breakpoints based on New York Stock Exchange (NYSE) issues only. All slocks are equally-weighted in a portfolio. The sample includes all NYSE, American Stock ExВ­change (AMEX), and Nasdaq domestic primary issues with coverage on the Center for Research in Security Prices (CRSP) and COMPUSTAT.

Panel A reports the average past six-month return for each portfolio, and buy-and-hold returns over periods following portfolio formation (in the followВ­ing six months and in the first, second, and third subsequent years). Panel Р’ reports accounting characteristics for each portfolio: book value of common equity relative to market value, and cash flow (earnings plus depreciation) relative to market value.

Panel РЎ reports each portfolio's most recent past and subsequent values of quarterly standardized unexpected earnings (the change in quarterly earnings per share from its value four quarters ago, divided by the standard deviation of unexpected earnings over the last eight quarters). Panel D reports abnormal returns around earnings announcement dates. Abnormal returns are relative to the equally-weighted market index and are cumulated from two days before the one day after the date of earnings announceВ­ment. In Panel E, averages of percentage revisions relative to the beginning-of-month stock price in monthly mean 1/B/E/S estimates of current fiscal-year earnings per share are reported.