PurposeThe study's purpose is to examine market returns around dividend announcements that contrast with a pattern of prior dividend announcements.Design/methodology/approachThe paper identifies firms that have a smooth dividend pattern of once-a-year dividend increases but at some point break that pattern and announce an unchanged dividend. The sample design allows the opportunity to investigate the market reaction to unchanged dividend announcements when an increase was likely to have been expected.FindingsThe results indicate that failing to increase the dividend is associated with significantly positive abnormal returns that are greater in magnitude for more entrenched dividend-increase records, supporting a contrast-effect hypothesis.Originality/valueThe results indicate that dividends are interpreted not only relative to the immediate dividend amount but also how the decision contrasts with dividends over a prolonged period. This finding suggests that the information content of the announcement of an unchanged dividend can vary according to the prior dividend pattern.
Read full abstract