Date of Award

Summer 1994

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Program/Concentration

Business Administration-Finance

Committee Director

Bruce Seifert

Committee Member

Edward Markowski

Committee Member

Tim McKee

Committee Member

Greg Noronha

Abstract

The tax effects hypothesis states that dividends have a negative impact on the value of a firm due to the preferential treatment given to capital gains over dividend income in some countries. This study tests the tax effects hypothesis in five countries: Australia, France, Germany, Japan and United States. The countries are selected because each had a significant tax law change within the period of study (1983-1991) and therefore provides an excellent opportunity for validation of this hypothesis.

The tax effects hypothesis is tested by first examining the effects of tax law changes on dividend payout ratios and then by studying the relationship between expected return and dividend yield before and after a tax law change in each country. A modified Capital Asset Pricing Model is used in examining this relationship.

Dividend payout ratios, dividend yields and dividend growth rates are calculated for each control to check if there are significant differences across countries. The final section of the study uses data in the five countries to test Lintner's partial adjustment model.

The results show that (1) There are some significant differences in payout ratios between the countries. The Australian and German firms have the highest dividend payout ratios while the French firms have the lowest dividend payout ratios. Australian firms also have the highest dividend yields and growth rates while Japanese firms have the lowest dividend yields and growth rates. (2) The post tax law dividend payout ratios of countries that increased the tax disadvantage of dividend income generally decreased. While the direction of the change in payout ratios supports the tax effects hypothesis, the amount of the change is insignificant in some cases. (3) A positive relationship between expected return and dividend yield is observed in countries that have higher effective tax rates on dividend income than capital gains. The relationship between expected return and dividend yield is positive and significant in France, Germany (after the tax law change), Japan and United States. An insignificant relationship is observed in countries that have similar tax rates on dividend income and capital gains. This is true for Australia and Germany (before the tax law change). These results suggest that dividends have a negative impact on the value of a firm and are consistent with the tax effects hypothesis. (4) Dividend behavior of firms in the countries can be explained by Lintner's model. The calculated payout ratios for Australia, France and U.S. are similar to the actual payout ratios. The calculated payout ratios for Germany and Japan are lower than the actual payout ratios but are within the calculated range of the payout ratios. Australia and Germany have the highest speed of adjustments among the five countries.

DOI

10.25777/a1ms-ac54

Share

COinS