Why Do Public Companies Begin Paying Dividends?
Posted: 25 Sep 1999
Date Written: November 1994
This paper examines empirically whether the dividend initiation decisions of a sample of newly-public firms are best explained by the predictions of the Miller and Rock (1985) cash flow signalling model, or by the competing agency cost models presented by Easterbrook (1984), Jensen (1986), and Rozeff (1982). Our results strongly support the agency cost models of dividend initiation. We document that dividend-initiating (DI) firms are older, larger, more profitable, and have higher levels of free cash flow at their initiation date than a matched set of non-dividend- initiating (NDI) companies. In the years after initiation, the DI firms issue equity and convertible securities more frequently, have higher levels of free cash flow, and have lower levels of--and greater post-IPO reductions in--insider shareholdings than do the matching NDI companies. These results suggest that firms begin paying dividends as a response to changes in the company's financial success and ownership structure.
JEL Classification: G14, G24, G35
Suggested Citation: Suggested Citation