Empirical investigation on the determinants of corporate dividend policy. Sectors considered: 'Food and drugs retailers' and 'Health care equipment and services'

Thèmes abordés

Dividends policy, statistical analysis

Résumé du document

The Corporate Dividend policy, also called the Dividend Puzzle (Black, 1976) has always beensubjected to a lot of studies since, at least, the paper of Modigliani and Miller in 1961. The work ofthese authors stated that, in a frictionless world, when the firm has a constant investment policy,its dividend payout policy will have no consequences on the wealth of the shareholder. Indeed, theirreasoning is that higher dividend payments lead to lower retained earnings and capital gains, andinversely, leaving the total wealth of the shareholders unchanged. But, as we can see in thenewspapers, corporations are following highly deliberate dividend payout policies. This is thusraises a question: How are firms choosing their dividend policies?
In order to answer to this question, we will first briefly study the history on the determinantsof the Corporate Dividend Policy and then conduct an investigation on this topic by analyzing twoparticular sectors of the London Stock Exchange: ‘Health care equipment and services' and‘Food and drugs retailers'.

Extraits

[...] That is what we were expecting to find because high debt rhymes with high fixed charges and thus lower dividends. Comparison of the two sectors When comparing the two sectors, we can see that they have different particularities. Indeed, the only two determinants that have the same sign are the profitability and the debt ratio. However, the profitability's coefficient is negative for both sectors and it is exactly the opposite of what we were expecting. Nevertheless, we can explain these negative results for the two sectors. [...]

[...] We were thus expecting a result much more negative for HCES than for FDR. When we look at the output of the regression for stability, we can see that the coefficient is about 6.59 for FDR and about - 0.67 for HCES. Thus the smoothing of the dividend payout has a really big and positive influence on the dividend payout for FDR and a poor and negative influence on HCES. Finally, concerning the size, both industries have a coefficient close to zero ( 0.12 for FDR and 0.12 for HCES). [...]

[...] We will look afterwards, when doing the regression, at their relationship with de dividend payout ratio (DPR) on the seven last years. To compute the DPR, we will divide the dividend per share (DPS) by the earning per share (EPS). The three main determinants explained by Fama and French (2001) have to be considered in this study, as the authors have proven that they are really significant. Moreover, other researchers have also used these determinants, for example Casey, Dickens and Newman (2002) and AlYahyaee, Pham and Walter (2008) Our first determinants will thus be the profitability of the firm. [...]

[...] We will first analyse the sector “Food and drugs retailers”, then the sector “Health care equipment and services”, and finally we will compare the two sectors in order to highlight the particularities of each one. Sector one: Food and drugs retailers Before looking at each factor, we can see on the output of the regression that for this sector, the adjusted R2 is at It means that the proportion of the variation in the dependent variable is well explained by the explanatory variables. [...]

[...] The signalling feature of dividend is subject to debate. Indeed, a lot of studies have shown that dividends convey information about the current or future level of earnings (Bhattacharya (1979), Miller and Rock (1985), John and Williams (1985)). But, on the contrary, other authors and notably Kumar (1988) showed that changes in dividends signal changes in the firm's perspectives, but that they are weak predictors of earnings. Dividends can also be useful for reducing the agency problem between managers and stockholders. [...]