세금과 자본구조가 이익-수익률 관계에 미치는 영향

초록

The expected return on equity is known as a theoretical determinant of the earnings-return relation, and it can be expressed as a function of leverage, corporate and investor level taxes. Dhaliwal et al.(2005) derive predictions relating leverage and taxes to earnings response coefficients and find that the coefficient linking earnings to returns in a reverse regression is increasing in leverage and that the effect of leverage on the return response is decreasing in a firm`s marginal tax rate. This finding is consistent with the tax benefit of debt, i.e. higher marginal tax rates, reducing the equity risk premium associated with leverage. On the other hand, they do not find conclusive evidence that investor level taxes affect the leverage-related component of the earnings-return relation. Differing from the United States tax system, in Korea, personal dividend income tax is somewhat mitigated due to tax relief through the provision of a dividend tax imputation credit system. Nevertheless, a personal investor`s dividend income is taxed at an effectively higher rate than capital gains income because capital gains income is not taxed except for the majority shareholders of KSE and KOSDAQ firms. Consequently, dividend tax penalty exists in the Korean tax system. The purpose of this paper is to investigate whether shareholder level taxes affect earnings response coefficients using KSE data. Because capital gains income and dividend income are subject to different tax rules between Korea and the United States, this paper can provide additional evidence for the prior results. For this purpose, I reexamine how financial leverage, corporate and investor level taxes affect the relation between earnings and stock returns. As an extension to the prior research, I add institutional ownership as a determinant of the earnings- return relation and examine how institutional ownership as well as leverage, corporate and investor level taxes affect the relation between earnings and stock returns. The tax disadvantage of dividends relative to the capital gains that exists for personal investors does not exist for most institutional and corporate investors. For example, dividends are taxed at the same effective rate as capital gains for tax-exempt institutional investors. Also, dividends are taxed at a lower effective rate than capital gains for corporate investors because corporations are entitled to a dividend-derive deduction. This suggests that if stock returns incorporate a dividend tax penalty, then the likelihood that low-tax shareholders alleviate this penalty rises as the level of institutional and corporate ownership increases. As a result, I expect that even though the personal tax penalty affects the firm`s earnings response coefficient, the magnitude of the effect is mitigated as the level of institutional and corporate ownership increases. In a forward regression specification, I find empirical evidence that the earnings response coefficient linking earnings to returns decreases in financial leverage, and that the effect of leverage on the earnings response coefficient decreases in a firm`s marginal tax rate, consistent with prior research. The results are consistent with the notion that the tax benefit of leverage increases in the marginal tax rate of the firm, which mitigates the effect of leverage on the earnings-return relation. In addition, I find empirical evidence that the personal tax disadvantage of debt increases the effect of leverage on the earnings-return relation. This is consistent with the interpretation that the personal tax disadvantage associated with debt increases the relative cost of debt compared to equity. Finally, I find evidence that institutional and corporate ownership which mitigates the personal tax disadvantage decreases the effect of leverage on the earnings-return relation as expected. These findings provide evidence on the leverage, taxes, and institutional ownership-related determinants of earnings response coefficients. Overall, these results are consistent with the notion that the tax penalty on dividends, relative to capital gains, reduces the earnings-return relation. Thus, the results suggest that shareholder income taxes influence equity value, and firm specific tax characteristics such as dividend policy and ownership structure affect how shareholder income taxes influence equity valuation. This study contributes to the literature by providing evidence of an additional factor, a firm`s owner structure, that may influence the earnings-return relation. It also contributes to the growing body of accounting literature that investigates whether and how taxes influence equity price.

키워드