Essays on Stock Return Predictability and Corporate Payout Policy
Garcia Ares, Pedro Angel
Thesis or dissertation
University of Exeter
This dissertation studies the effect of the changes in corporate dividend policy on the predictability of stock returns. The first two chapters re-visit the question of what drives stock returns after controlling for these market-wide changes in the cross-sectional profile of dividend paying firms, and the third chapter studies the nature of these changes across different industry groups, stock market indexes, size and age. There have been several significant changes in the nature of corporate payout policy of US firms over the last several decades. We focus our work on the effect of two of these changes, using the present value model, into the question of what drives stock returns-cash flows or discount rate news. Chapter 1 studies the effect of the large decrease in the proportion of dividend paying firms in changes in expected cash flows and/or discount rates by focusing on portfolios of dividend paying firms rather than aggregate portfolios of all listed firms. Our results, from Chapter 1, imply that the relatively importance of cash flows and discount rate news is intimately related to the cross-sectional variation in the patterns of dividend payers in the stock market and provide an intuitive explanation for the contradictory results documented in the existing literature. Chapter 2 builds on Chapter 1 and tries to reconcile and explain why results using the return and the book-to-market decomposition differ for the post-WW II period. It also provides with an alternate explanation different from dividend smoothing for the apparent absence of dividend growth predictability in post-WW II U.S. data. We find that predictive regressions based on the return on equity decomposition are sensitive to the way in which firm-level data is aggregated. Specifically we find that when firm-level data is weighted by value both decomposition methods -the Campbell-Shiller and the Voulteenaho return decompositions provide strong support for cash flow news as a driver of stock returns in post-WW II data. We also find that, in post-WW II data, the existence of cash flow news is driven by the fact that the biggest firms by market capitalization are not always those that generate the biggest earnings or pay the largest dollar dividends. In Chapter 3, the final part of this work, we investigate the anatomy of corporate payouts. Specifically, we use firm-level data to understand which firms drive the changing patterns of payouts over time and in the cross-section. Our work extends the current literature by studying firms payouts based on industry sectors, firm age and other attributes. Our main finding is that we find support for our conjecture - in Chapter 2, that the biggest firms by market capitalization are not always those that generate the biggest earnings or pay the largest dollar dividends.
PhD in Finance