Merton Miller and Modern Finance

Financial Management (Financial Management Association),  Winter, 2000  by Rene M. Stulz

• E-mail
• Print
• Link

Merton Miller was always skeptical that the bondholder-shareholder conflict or the underinvestment problem could explain why firms did not take greater advantage of the tax shield of debt. Not surprisingly, his skepticism was the result of the role of arbitrage in how he thought about economic phenomena. If the tax shield of debt was so large, why was it that investment bankers would not devise solutions that would enable firms to take advantage of this tax shield and overcome the costs of debt through clever contracting? As always, he wanted a solution to the problem that would not provide clever arbitrageurs with profit opportunities. He pointed out that for bankruptcy costs, debt securities that would avoid these costs already exist and that if these costs were large, one would expect these debt securities to be used much more than they are. With income bonds, interest has to be paid in any year only if earned, but it is fully tax deductible if paid. This made it hard for Merton Miller to believe that dir ect bankruptcy costs or the impact of high leverage on investment were the explanation for why firms did not have more debt in the presence of an apparently large tax advantage to debt.

In 1977, Merton Miller revisited the issue of the impact of corporate taxation on the irrelevance propositions in a classic paper titled "Debt and Taxes" that shows perhaps better than any of his other papers how he could use arbitrage arguments to change how finance academics and practitioners understood how the world works. In that paper, he pointed out that the tax advantage of corporate debt might be mostly if not completely illusory. Because interest on corporate debt is taxed as income for the holder of corporate debt, the interest paid on corporate debt must be high enough so that the after-tax income from holding corporate bonds is attractive relative to the income from equity which, when it accrues as capital gains, is taxed at a lower effective rate. As a result, corporations get to deduct from their taxes interest payments but, because personal taxes on interest income are higher than on capital gains, the before-tax cost of capital on debt must be higher than on equity if investors are to hold de bt.

In his paper, Merton Miller showed that under specific conditions the leverage irrelevance proposition could hold even in the presence of taxes. In his argument, the tax rate on personal income from bonds is progressive, but there is no income tax on equity income. Bonds are risk-free. There is a demand for bonds which is satisfied by taxable and tax-exempt bonds. Merton Miller shows that in equilibrium the coupon on taxable bonds must be the coupon on tax-exempt bonds grossed-up for taxes at the corporate tax rate. Let [r.sub.0] be the tax-exempt rate and [[tau].sub.c] be the corporate income tax rate. With this notation, the interest rate on corporate bonds is [r.sub.0]/(1-[[tau].sub.c]). If the interest rate on corporate bonds is say [r.sub.c] [greater than][r.sub.0]/(1-[[tau].sub.c]), corporations prefer equity financing since the risk-adjusted cost of equity is [r.sub.0] while the cost of debt after tax is (1-[[tau].sub.c])[r.sub.c][greater than][r.sub.0] Alternatively, if the interest rate on corporate bonds is lower than [r.sub.0]/(1-[[tau].sub.c]), corporations would rather obtain debt financing than equity financing since the after-tax cost of debt is lower than the risk-adjusted cost of equity. As a result, the only feasible equilibrium is the one where the after-tax cost of debt is exactly equal to the after-tax cost of equity. When this equilibrium obtains, no firm has a financial incentive to alter its mix of debt and equity even though interest payments on debt are tax deductible. The "Debt and Taxes" paper led to a large literature investigating whether the leverage irrelevance condition holds in the presence of personal and corporate income taxes. The paper also made clear that the perfect markets assumptions are sufficient conditions for leverage to be irrelevant, but not necessary conditions. Saying that the assumptions required for Proposition I do not hold is not enough to conclude that leverage matters. For leverage to matter for the value of a corporation, it has to be the case that no clev er arbitrageurs can make money from such a situation.