Dividend and Capital Gains Taxation in Firm Valuation: New Evidence
Posted: 11 Aug 1997
Date Written: April 1997
Abstract
This paper considers how personal taxation of dividends and capital gains influences the valuation of a firm. Although this basic issue has been debated for more than twenty years in the corporate finance, public finance, and accounting literatures, the impact of personal taxation upon firm valuation and the cost of equity capital remains controversial. Most valuation studies use a dividend discount model as the starting point. This includes the residual-income valuation model which is used increasingly to justify price-based association studies. However, the formal models and the current empirical research largely ignore dividend and capital gains taxes. Since corporate returns on investment generate retained earnings and are subject to dividend taxes upon distribution, and returns of capital stock are not, basic intuition suggests it is likely that retained earnings and capital stock are valued differently. This paper demonstrates how the residual income model can be adapted to incorporate taxes on dividends and capital gains and empirically tests whether retained earnings is priced differently than capital stock in the manner suggested by intuition and the model. In addition, the study illustrates how personal taxes influence the persistence of residual income, which affects the association between price and earnings. The hypotheses developed in the paper abstract from tax issues relating to debt versus equity, instead focusing on how personal taxation impacts retained earnings versus (external) capital stock equity financing. Hence, the research contributes to the controversial ongoing debate in the economics and finance literatures revolving around the traditional versus new views of dividend taxation. Under the traditional view, it is presumed that future dividend and capital gains taxes are not capitalized into prices. In contrast to the traditional view, the new view posits that as long as retained earnings are expected to be realized through dividend distributions, future dividend taxes are capitalized into prices. Previous attempts to empirically examine dividend tax capitalization have led to mixed and somewhat inconclusive results, prompting Zodrow [1991] to state that "this question is still one of the most controversial ones in public finance." The tests in this paper provide an alternative perspective which overcomes many of the research design difficulties hampering the prior research. Empirical results provide robust support for a valuation model that incorporates personal taxation on dividends and capital gains. In particular, results indicate that the relative valuation weights investors assign to shareholders' equity versus earnings critically depend on firm-level ratios of retained earnings to total equity. An extensive set of sensitivity tests is used to control for alternative, non-tax explanations for the findings. A key implication of the results is that the ratio of retained earnings to total equity represents a potentially important omitted variable for many prior studies evaluating associations between price and earnings or equity. The results also suggest the cost of retained earnings is generally lower than the cost of paid-in equity capital, and the value of a dollar of retained earnings is less than the value of a dollar of paid-in equity capital. From a tax-policy perspective, the empirical results add evidence in favor of the new view of dividend taxation.
JEL Classification: G12, G14, G32, G35
Suggested Citation: Suggested Citation