Interaction of Stock Return Momentum with Earnings Measures

Examination of the interaction of stock return momentum with various earnings measures finds that large-capitalization companies with poor past returns and high return on equity (ROE) significantly underperform the market and companies with poor past returns and low ROE. Thus, the profitability of high-ROE companies with poor past returns may have peaked. In addition, companies with poor past returns and poor earnings quality (as measured by accruals) significantly underperform the market and companies with poor past returns and good earnings quality. Therefore, the market may not fully recognize manipulation of earnings. The findings are consistent with the explanation that momentum is driven by slow reaction to news.This article examines the interaction of stock return momentum with various measures of earnings. I had two goals in mind. First, such an analysis could help investors and portfolio managers understand the dynamics of the stocks in their portfolios and thus enhance portfolio performance. Second, and as important, the analysis deepens the profession’s theoretical understanding of the intermediate-term momentum phenomenon.I computed the historical performance of 25 bivariate stock quintile portfolios. Stocks were placed in the first set of quintiles based on their past year’s return and in a second set of quintiles based on an earnings measure. The universe is the S&P 500 Index, and the historical time period is 1970 to 2004. The earnings measures that I report are trailing return on equity, change in ROE, and earnings quality (as measured by balance sheet accruals). I also used as earnings measures forecast ROE, change in forecast ROE, P/E, and price-to-book ratio (results reported in the appendix). Results for a one-month holding period are reported (with results for a six-month holding period available online).I found that large-cap companies with poor past returns and high ROEs not only significantly underperform the market but also underperform companies with poor past returns and low ROEs. (Similar results were found for forecast ROE, change in ROE, and change in forecast ROE.) This finding suggests that the profitability of companies with high ROEs and poor past returns may have already peaked. In addition, I found that companies with poor past returns and poor earnings quality significantly underperform the market and companies with poor past returns and good earnings quality. This finding implies that company managers’ manipulation of earnings may not be fully recognized by the market, which enhances the momentum effect. These findings are consistent with the behavioral explanation that the momentum phenomenon is driven by the market’s slow reaction to news.