Merton Miller and Modern Finance

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

If the only departure from the assumptions leading to Proposition I is a tax subsidy to corporate debt, one would expect to observe extremely high leverage since extreme leverage would maximize the value of the tax subsidy. Empirically, however, leverage is not extreme. In fact, some firms even have negative leverage--they pay more in taxes on securities held than they get to deduct because of paying interest on debt. To make sense of the limited levels of leverage in the presence of what appeared to be a large tax subsidy for debt, finance had to either relax other assumptions leading to Proposition I or conclude that the subsidy was illusory. Initially, the route chosen by finance was to take into account bankruptcy costs. Bankruptcy costs are effectively the outcome of contracting costs--as the firm defaults, the firm cannot be costlessly reorganized. In the presence of bankruptcy costs and tax benefits of debt, each firm has an optimal amount of debt such that the increase in the present value of expected bankruptcy costs resulting from an additional dollar of debt for that firm is exactly equal to the present value of the expected tax benefits from that additional dollar of debt.

Merton Miller was skeptical that expected bankruptcy costs could be large enough to explain why firms did not take advantage of the tax subsidy of debt more. His assessment of the evidence on bankruptcy and financial distress costs was that "neither empirical research nor simple common sense could convincingly sustain these presumed costs of bankruptcy as a sufficient, or even as a major, reason for the failure of so many large, well-managed US corporations to pick up what seemed to be billions upon billions of dollars of potential tax subsidies" (Miller (1991a), p. 274). This assessment led him to one of his most memorable statements, namely that "the supposed trade-off between tax gains and bankruptcy costs looks suspiciously like the recipe for the fabled horse-and-rabbit stew--one horse and one rabbit" (Miller (1977), p. 264).

Since direct bankruptcy costs could not explain why firms were not taking advantage of the apparent tax subsidy of debt, finance turned to other explanations for low leverage based on contracting costs. Jensen and Meckling (1976) showed that, as leverage increases, shareholders become tempted to take advantage of bondholders by increasing the volatility of the firm even if doing so would hurt shareholders if the firm had less leverage. The bondholder-shareholder conflict identified by Jensen and Meckling makes debt more costly because it forces firms to behave inefficiently as a result of leverage or to spend real resources to insure that they will not take advantage of bondholders. Myers (1977) showed that when firms are highly levered, the existing shareholders may choose not to issue equity to finance projects because the new equity will increase the value of the firm's debt at the expense of the value of the equity of the existing shareholders. Myers called this problem the underinvestment problem and ide ntified it as a cost of leverage because firms with high leverage invest less than firms with lower leverage. In a world of perfect markets, neither the underinvestment problem nor the bondholder-shareholder conflict would exist because in a world without contracting costs, the capital structure of the firm can always costlessly be changed to make sure that a firm can take advantage of all profitable investment opportunities and does not invest inefficiently as a result of leverage.