Do dividends announcements signal future earnings changes for Jordanian firms?

Abstract

Purpose This study aims to examine the signalling hypothesis of dividends by testing empirically the market reaction to dividends announcements. Furthermore, this study aims to examine the information content of dividends announcements with respect to future earnings changes for a sample of Jordanian industrial firms over the period 2009 to 2015. Design/methodology/approach The authors mainly used the event study methodology to examine the market reaction to dividend release announcements. The market model is used to generate the expected returns. Also, the t-test is used to examine the significance of the mean and cumulative abnormal return. Furthermore, a simultaneous-equation model developed by Nissim and Ziv (2001) and Grullon et al. (2005), applying the two-stage least squares (2SLS), is used to examine the relationship between dividends changes and future earnings changes. Findings The results reveal consistency with the limited extant empirical evidence for developing markets and provide some new insights for Jordanian listed firms that support the signalling hypothesis. In applying the event study methodology, the information content of dividends shows that there is a significant positive market reaction to dividends announcements. The study’s findings also present a strong relationship between dividends announcements and profitability in the year of announcements and the subsequent year, whereas this relationship does not exist in the second year. The findings show that there is value-relevance for dividends, suggest that investors recognize the signalling purpose and discern that dividends announcements are useful in predicting favourable and unfavourable future earnings in the short run (the same year and subsequent year) and also show that managers may use dividends to signal earnings prospects in anticipation of expected future market benefits. Research limitations/implications The findings of this study could have significant policy implications. The support of a signalling effect implies an existence of information symmetry, at least theoretically, between management and investors. On the other side, this study could not reflect the levels of inside ownership or the existence of signalling substitutes even though these findings could have implications for Jordan’s existing corporate governance practices and firms’ disclosure environment. The results are specific to Jordan, but they do shed light on the generality of the rival models of dividend policy. Many of the structural characteristics of the capital market in Jordan are, however, also present in other emerging markets. The results from this study may, therefore, help provide the basis for comparative research both in the region and in other emerging markets. Practical implications The support of the signalling effect implies the existence of information symmetries, at least theoretically, between management and investors. These findings could have implications for Jordan’s existing corporate governance practices and firms’ disclosure environment. Originality/value This paper contributes to the literature by providing a workable test for the dividend signalling hypothesis, applying a simultaneous-equation model that incorporates the market reaction to dividends announcements and future earnings changes. Moreover, this paper uses a recent data set of dividends announcements in Jordan. This study provides additional insight to support the signalling hypothesis in emerging markets. Overall, current and previous studies have focused typically on investigating dividend policy in developed markets, especially the US and European markets, although there has been limited analysis of dividends changes on earnings changes for developing markets.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 31 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.