Why Do Financially Distressed Firms Increase (or Initiate) Dividends?

Abstract

This study focuses on the short- and long-term performance of financially distressed firms that suffer a negative net income in the observation quarter but consequently decide to change their dividend policy by increasing (or initiating) dividends. With the quarterly panel data over the period 1962-2006, we first document a general reversal pattern in financial performance: firms that suffer a negative income quarter tend to improve their performance to different degrees during the next quarter across all firm sizes, no matter what dividend policies those firms adopt. After further analysis, we find strong evidence that firms choosing to increase (or initiate) dividends after suffering a negative net income quarter exhibit significantly better financial performance in net income, ROE, and ROA, after changing their dividend practice. After removing the possible autocorrelation and seasonality (including the earnings reversal pattern) in earnings per share series, a regression model confirms that adding the change in dividend per share during the observation quarter predicts the change in earning per share in the follow-up quarters, evidenced by highly significant regression coefficients and improved mode fitness. Different from the previous findings that dividend changes don't signal future earnings growth, our results reveal information content in the change of dividend policy by financially distressed firms and are consistent with signaling and dividend smoothing hypotheses. Sorting firms by their enterprise values (a proxy for size), we further find that, except for the firms in the smallest quintile, this information content is strong, reliable, and persistent across all the other financially distressed firms that decide to change their dividend policies. This effect seems to last not only for the short term (up to 1 year) but also for the long term (up to 5 years).