An empirical analysis of stock returns around dividend changes

This paper documents significant drift in stock returns following announcements of changes in cash dividends. The magnitude is (i) smaller for increases than for decreases, (ii) inversely related to firm size and positively to dividend yield change, and (iii) concentrated in the first quarter. Beta changes do not explain the drift and it is robust in various subperiods. Next it is shown that dividend increases are positively autocorrelated especially every fourth quarter. The prices keep reacting to future announcements as if the market ignores these autocorrelations. Dividend decreases exhibit weak autocorrelation and the returns are negative for the following three announcements.