Predictable dividends - Empirical support for the predictability of quarterly dividend increases using accounting data, financial distress predictions and smoothing behaviour
Abstract: We investigate whether, and show support for that, quarterly dividend increases can be predicted using accounting data combined with smoothing behaviour and financial distress predictions. The empirical findings are achieved through the development of a probabilistic model for dividend increase prediction with data for U.S. manufacturers from year 2000 to year 2016. However, there is reason to doubt the general applicability of the model, as differences are shown both over time and across sub-industries. Despite positive initial results, the financial distress prediction variables are not clearly beneficial to the model. When using the proposed dividend increase prediction model with a suggested cut-off probability for our sample, a Naïve model is outperformed and the results improve further for our separate validation sample. A strategy to purchase the shares of the companies for which dividend increases are predicted achieve abnormal returns which are significant at the 0.10 level.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)