배당소득세불이익과 기관지분율이 내재자본비용에 미치는 영향

초록

Whether stock prices capitalize shareholder level dividend taxes has been long question, but still there is no consensus among researchers. This question is important because it provides information about the determinants of a firm`s cost of equity capital. Recent accounting researches, specially in return studies, find evidence consistent with dividend tax capitalization using longterm and short-term returns. But, as Dhaliwal et al.(2005) pointed out, one important limitation in these studies is that they use realized returns to examine the effect of dividend taxes on stock returns because ex ante required returns are not readily avaliable. This paper extends the prior return research by estimating the implied cost of capital using Korean market data and investigate the effect of dividend taxes on these measures of the ex ante cost of equity capital to test for dividend tax capitalization. The ex ante cost of equity capital is recognized as the theoretically relevant measure in tests for dividend tax capitalization. As for the implied cost of capital, we estimate 3 different measures using the methods established by Gode and Mohanram(2003) and Easton(2004). Korea is different from the United States in that Korea has dividend imputation system to protect shareholders from double taxation whereas the United States has not. As a result, Korean market shareholders faces different dividend tax penalty than the United States market shareholders. Our use of the implied cost of capital and the Korean market data should provide an international evidence of dividend tax capitalization. We also use institutional ownership to further investigate the question：if dividend taxes affect the cost of capital, how does the magnitude of this effect vary with this firm specific characteristics? We limit the sample period for 2 years from 2001 to 2002 considering the estimation of implied cost of capital and the measurement of dividend tax penalty and use total 281 firm-year data. We use the methodology in Dhaliwal et al.(2005) by exploiting time-series variation in dividend tax penalty, which is a function of dividend and capital gain tax rates, to isolate the tax effect from the information and agency effects of dividends. A firm`s dividend yield is used to capture the level of tax disadvantaged dividend income and the level of institutional ownership is utilized as a proxy for marginal investors which have relatively low dividend tax penalty. We further disaggregate institutional ownership into 3 different groups that are more homogeneous with respect to their tax attributes：government, banking-securities-insurance company, and other corporations` ownership ratios. The results generally support the dividend tax capitalization hypotheses. First, the relation between the implied cost of capital and dividend yield is increasing in the magnitude of dividend tax penalty. Second, the positive relation between the implied cost of capital and dividend yield is decreasing in the level of institutional ownership. Third, the positive relation between the implied cost of capital and dividend yield is decreasing in the level of banking-securities-insurance company, and other corporations` ownership ratios when we further disaggregate institutional ownership into 3 different groups. This result suggest that banking-securities-insurance company, and other firms can play their role as marginal investors in Korea. Overall, these results suggest that dividend tax penalty increases the dividend tax premium, in other words implied equity premium, and accordingly it has negative effect on stock prices. The evidence in this paper adds to the understanding of the effect of taxes on equity prices.

키워드