這篇論文主要在探討以不同融資方式為主的經濟體，是否會對個別公司的股利政策造成影響，我們針對15個國家1244家公司作探討，發現在以銀行債為主要融資方式的國家，個別公司相對發放較低的股利，並且現金股利亦呈現較高的波動度，顯示資訊不對稱的嚴重性，是造成不同資本市場有不同股利政策的原因之一。 / Abstract This paper examines whether differences in the sophistication of capital markets drive dividend policies to vary across countries, by studying on dividend policies of 1244 firms in 15 countries. We find that reliance on banking sector causes lower dividend payout and incentives for dividend smoothing. The same relationship is also found between quality of accounting system and dividend policies. These results confirm that less information asymmetry defuses the need for using dividends as signaling and discipline tools. Furthermore, we find that there exists a positive relation between development of stock market and payout level. This indicates that more important the stock in providing capital to firms, the more likely the firms increase their dividend payout. Finally, some evidences support that a shorter investment horizon leads to a higher payout ratio and a higher extent of dividend smoothing.

We propose an asset pricing model in a production economy where cash flows are determined by firms' optimal dividend and investment decisions. Extensive and intensive decision margins in dividend payout are modeled with cash holding and investment adjustment costs. The model implies that delays in dividend distribution of young and growing firms play instrumental roles in explaining various asset pricing anomalies. Quantitative results show that model-implied dividend policies and investments are consistent with data, and the cross sections of stock returns are well explained by the interactions between productivity shocks and the lumpy dividend policies. Additionally, the model produces countercyclical variations in the market risk premium. In addition, we empirically investigate the relevance of firm characteristics and aggregate productivity shocks in determining dividend payment propensity, thereby asset prices. It is found that excess returns for dividend payers over nonpayers are significantly linked to business cycles. Relative future returns are fairly predicted by the spread of lagged propensities to pay dividends. Furthermore, the empirical results document that each future return of payers and nonpayers increases in propensities to pay out cash to shareholders. These results are consistent to our rational expectations model of dividend policy, and contradictory to the catering theory of dividends.

The determinants of corporate dividend policy remain controversial despite half a century of active research. Over that time a number of competing theories of dividend policy have been proposed, but no consensus has been reached about their explanatory power. This thesis examines the determinants of dividend policy of publicly quoted companies in Jordan as a case study of an emerging market. The study uses a firm-level panel data set of all publicly traded firms on the Ammam Stock Exchange between 1989 and 2000. Nine research hypotheses are developed, which are used to represent the main theories of corporate dividends. The results of studies conducted in this thesis suggest that the proportion of stocks held by insiders and state ownership significantly affect the amount of dividends paid, but not the decision to pay dividends. Larger, mature, profitable firms with less investment opportunities are more likely to pay dividends. These factors are found to also positively affect the level of dividends. Results provide no support for the signalling hypothesis. The thesis concludes with a discussion of some of the implications of all results and suggestions for further research. / Doctor of Philosophy (Finance)

NONE

The number of domestic and cross border bank mergers and acquisitions (M&A) has increased over the last decade with a resultant impact on the bank dividend. This study examines the effect of M&A on the dividend policy by comparing the abnormal returns, profitability and dividend policy of the domestic and cross border bank acquirers. The study focused on EU mega-bank mergers and acquisitions within 1997-2007 involving only commercial-to-commercial banks. The sample consists of a total of 62 mega-M&A with a minimum deal value of €500 million. Three hypotheses were formulated specifically to test: (i) the wealth effect and geographical diversification of the M&A between domestic and cross border acquirers; (ii) the effect on in the financial performance of both acquirers and (iii) the M&A impact on dividend policy on banks after bank M&A. Two strands of the literature were reviewed focusing on M&A and dividend policy. The event study methodology was used to calculate the abnormal returns of both the domestic and cross border acquirers which were standardised. A long window of 61 days was applied to capture a satisfactory length of pre and post merger events that could capture the behaviour of the abnormal returns and consequent effect on dividend policy. The hierarchical regression model was used to estimate the impact of the variables on the profitability and dividend policy of the acquirer banks. In comparison with the domestic acquirers, the cross border abnormal returns showed a trend of significant negative results following the M&A announcement. The domestic acquirers showed no significance but, on average have higher cumulative total standardised abnormal returns (CTSAR) than the cross border acquired banks. The result of the financial performance showed that CTSAR of the cross border acquirers is significantly affected by the profitability of the banks but insignificant with domestic acquirers. However, the cost-to –income ratio (CIR) significantly affects the performance of both bank acquirers. CIR and RISK (measured by the ratio of the loan provision to net interest revenue of the banks) highly correlated with profitability of both the domestic and cross border acquirers. The management of costs and loans risks were found to be significant variables in the achievement of profitability among domestic acquirers. The dividend policy hypothesis result indicated that CTSAR has a weak correlation and insignificant effect on the dividend policy variables. Infact, the Causality test result confirmed that the CTSAR does not Granger cause dividend policy. However, the study provides strong support to previous studies that beta, liquidity, taxes, and the finance structure of the acquirers are significant variables in the formulation of the dividend policy of the merged banks. The beta, which a proxy for risk, is the most significant factor affecting the dividend policy of the merged banks.

Abstract¡G From previous studies we found that growth opportunity is an important factor help explaining the cross-sectional differences of firms¡¦ dividend and financing policies. However, growth opportunities are unobservable and not measurable, scholars used different proxies to catch the idea. Prior empirical studies used market-to book, earning-price ratio, Tobin¡¦s Q, R&D intensity, capital expenditure (deflated by book value of assets), PPE, previous sales growth as proxies for growth opportunities, but had inconsistent results. The purposes of this thesis are: first, to find out the relationship between growth opportunities and realized growth, and to determine which proxy is most suitable; second, to find the empirical evidences of the relation between growth opportunities and financing policies, growth opportunities and dividend policies support contract theories, tax-based theories or signaling theories; and finally, to find out if firm¡¦s growth opportunities are affected by its financing and dividend policies. We examine all nonfinancial firms listed on the Taiwan Stock Exchange from 1991 to2000. We use the correlation and regression to determine the relation between growth opportunities and realized growth, then distinguish total debt ratio, short-term debt ratio, long-term debt ratio from financing policies, and dividend payout ratio, cash-dividend yield, stock dividend yield, and dividend yield from dividend policies. We use different growth opportunity proxy to separate whole sample into two subsamples, and growth as a dummy variable. We use regression to find the relation between growth and financing policies, dividend policies. Finally, we use simultaneous equation to determine the relationship among growth opportunities, financing policies and dividend policies to see if they are interdependent. Our empirical result shows that growth opportunities and realized growth have positive relations. But growth opportunities are not positively correlated with realized investment growth. Price-based proxies are often overestimate future equity market value growth. Growth firms have higher debt ratios and long-term debt ratios. The result supports contracting theory and progressive tax rate theory. Growth firms have higher short-term debt ratios, too. We can¡¦t find the significant relation between growth opportunities and dividend pay¡Vout ratios. Growth firms have lower cash dividend yield, and the result supports cash-flow constraint and contracting theories. There is no clear relation between stock dividend yield and growth opportunities. Through simultaneous equation, over financing and too much cash-dividend restrict firms¡¦ growth opportunities.

No description available.

No description available.

Thesis advisor: Edith Hotchkiss / Chapter one of this dissertation provides new evidence on the existence of dividend clienteles for institutional investors. We directly examine individual institutions' preferences for dividend paying stocks based on the characteristics of stocks held in their portfolio. Many institutions follow persistent investment styles, maintaining relatively high or low dividend yield portfolios over time. Institutions which hold portfolios of higher yielding stocks are significantly more likely to increase their holdings in response to a dividend increase or sell their stock in response to a decrease. For a subset of institutions, we directly observe the proportion of their portfolio managed on behalf of taxable clients. Consistent with tax-induced dividend clienteles, institutions with more taxable clients are less likely to increase their holdings in response to a dividend increase. Finally, we show that stock price reactions to announcements of dividend increases are related to characteristics of the institutions holding the stock. Our results suggest that tax status, as well as other factors are important in explaining observed clientele behavior. Chapter two explores the determinants of heterogeneity in institutional investor portfolio preferences and the relationship between institutions and the clients they serve. I find that the characteristics of an institution's clients and the characteristics of the institution itself are both important determinants of portfolio preferences and trading behavior. Specifically, I find that institutions traditionally subject to prudent investor laws are more likely to invest in high quality stocks, although, institutions sub-managing money for pension funds are less prudent than pension managers themselves. In addition, I find that institutions with taxable clients are likely to avoid unnecessary dividend taxation and turn over their portfolios less frequently. More generally, institutions exhibit systematic shifts in their exposure to common risk factors that may be explained in part by the levels and changes in client composition. While evidence for a causal link between client shifts and institutional preferences is limited to mutual funds, contemporaneous changes in clients and portfolio characteristics suggest that the dynamics of institutional investment are closely related to the nature of the clients served. / Thesis (PhD) — Boston College, 2007. / Submitted to: Boston College. Carroll School of Management. / Discipline: Finance.

Este estudo teve como objetivo obter evidências sobre a relação entre a política de dividendos adotada pelas empresas listadas na BM&FBOVESPA e as suas respectivas estruturas de propriedade e de controle. Adicionalmente, buscou-se identificar possíveis alterações ocorridas no tempo no que se refere às estruturas de propriedade e de controle e aos níveis de dividendos pagos pelas companhias. Para tanto, foram analisadas 297 companhias abertas no período de 2005 a 2012, cujos dados estavam disponíveis no banco de dados Economática, através da aplicação do modelo Tobit. Os resultados apontam que empresas com estruturas de controle e de propriedade concentradas tendem a distribuir menos dividendos. Também foi constatado que houve redução nos níveis de alavancagem de controle (diferença entre concentração de controle e de propriedade) nos últimos anos, mas isso não resultou em níveis mais altos de distribuição de dividendos. De modo geral, os resultados confirmam a existência de conflitos entre controladores e minoritários no Brasil. / This study aimed to obtain evidence on the relationship between the dividend policy adopted by companies listed on BM&FBOVESPA and their respective ownership and control. In addition, the study sought to identify possible changes in time with respect to the ownership and control and the levels of dividends paid by the companies. For this, 297 public companies were analyzed in the period from 2005 to 2012, using data available in Economática and applying the Tobit model. The findings point that firms with concentrated ownership and control tend to distribute less dividends. It was also found that there was a reduction in leverage levels of control (difference between concentration of ownership and control) in recent years, but this has not resulted in higher levels of dividend payments. Overall, the results confirm the existence of conflicts between controlling and minority shareholders in Brazil.