Deen Kemsley

Valuation of the Debt-Tax Shield

In this study, we use cross-sectional regressions to estimate the value of the debt-tax shield. Recognizing that debt is correlated with the value of operations along nontax dimensions, we estimate reverse regressions in which we regress future profitability on firm value and debt rather than regressing firm value on debt and profitability. Reversing the regressions mitigates bias and facilitates the use of market information to control for differences in risk and expected growth. Our estimated value for the debt-tax shield is approximately 40 percent (ten percent) of debt balances (firm value), net of the personal tax disadvantage of debt. In addition, our estimates for the debt-tax shield vary across tax regimes and across firms in a predictable manner.

Dividend taxation in firm valuation: New evidence

In this paper we develop a residual-income model showing how taxes on dividends affect the relative valuation of retained earnings versus contributed equity, as well as the value of expected future earnings. Tests of predictions from our model for a sample of Compustat firms from 197594 suggest that overall firm value, and the relative valuation weights investors assign to retained earnings, contributed equity, and current earnings, all critically depend on dividend taxes. The findings also suggest that investors take a proprietary perspective in valuation and impute an unrecorded shareholder-level tax liability on retained earnings. The U.S. tax system subjects retained earnings to dividend taxes upon distribution to shareholders, while contributed equity is returned to shareholders as a nontaxable return of capital. From the shareholders (proprietary) perspective, therefore, retained earnings should be valued on

The composite dividend tax rate

Dividends often impose taxes on investors. However, as certain prior financial models indicate, they also can produce a tax gain from leverage. Hence the composite marginal dividend tax rate can be specified as the nominal rate minus the offsetting tax gain from leverage. Although this principle has been embedded in theoretical models for more than 40 years, no prior study has examined empirically whether the dividend-induced tax gain from leverage influences dividend policy. We address this empirical void and find dividends decrease in the nominal dividend tax rate and increase in the offsetting tax gain from leverage. In addition, we find the composite tax rate outperforms traditional measures in explaining dividend policy for our full sample of firms. Consistent with prior theory, we also find the composite rate varies in influence according to the financing source for a dividend.