<p>Thaler’s compound prospect theory hypothesizes that gains should be segregated, and losses should be integrated to maximize an individual’s value function. However, these hypotheses have not been empirically tested in the context of corporate finance. To test this theory, we utilize a rich dataset of dividend changes. We find that the announcement returns are significantly higher when firms implement two smaller, sequential dividend increases rather than a single equivalent dividend increase. Conversely, the announcement returns are significantly more negative when firms segregate losses into two sequential dividend cuts rather than a single dividend cut of the same total amount. All analyses control for the propensity to make sequential dividend changes. These findings provide robust empirical evidence in corporate finance that supports Thaler’s compound prospect theory regarding gain segregation and loss integration.</p>