One of the great successes of modern contract theory is the costly state verification (CSV) environment, developed originally by Townsend (1979), and subsequently extended by Diamond (1984), Gale and Hellwig (1985), Williamson (1986, 1987), and many others. Under the assumptions of risk neutral agents and fixed verification costs, this environment delivers standard debt contracts as an optimal contractual form. In addition, it delivers a well-defined optimal capital structure for a particular sort of firm. More specifically, internal funds, provided by an informed entrepreneur, are exchanged for a residual claim, which resembles "equity." External funds, provided by lenders who must monitor (at cost) to become informed, are exchanged for contracts that promise payments which are a piece-wise linear function of firm performance. These claims resemble "debt." In this environment—and in some more complicated ones such as those studied by Boyd and Prescott (1986), Bernanke and Gertler (1989), or Boyd and Smith (1993)—the firm is optimally financed with this combination of "debt" and "equity."
This kind of model, where "equity" is owned entirely by insiders (say management), and where all external funding takes the form of debt, is a reasonable description of a very limited class of firms, however—for example sole proprietorships or other small owner-managed firms. Moreover, as discussed by Bernanke and Gertler (1989), the equity investment decision is rather trivial, since it is always optimal for firm management to invest its entire wealth endowment in the firm.
This sort of capital structure does not look much like that employed by large corporations in the United States or most other developed nations. In particular, major corporations typically have a large class of outside equity investors who are not necessarily any better informed than are debt holders. In most cases, there are also inside owner-managers and directors who hold equity and who, presumably, are better informed than outsiders.
In this paper, we attempt to provide a formal analysis of this kind of capital structure using the same contract theoretic approach that has proved so successful in understanding the use of debt contracts. It turns out that it is possible to do so with only a fairly slight modification of the standard CSV environment. In particular, instead of presenting entrepreneurs with only a single class of investment opportunity, we allow entrepreneurs to make use of convex combinations of two classes of investments. The first is exactly like that available in the standard CSV setup. Return realizations are costlessly observed only by an entrepreneur, and all other agents must pay a fixed monitoring cost to become informed about return realizations. The second (and new) type of stochastic investment opportunity is one in which return realizations are costlessly observed by all agents. For obvious reasons, we assume that the expected rate of return on the former exceeds that on the latter.
This minor modification of the underlying environment results in a considerable difference in the optimal financial structure of the firm. In order to characterize it, we begin by solving an optimization problem that determines each firm's investment in each type of project (that is, we determine its asset structure), and that determines the aggregate payments made by the firm as a function of the returns it receives. We then show that an optimal contractual arrangement can be supported by having the firm issue an appropriately chosen amount of debt (of the standard CSV variety) and an appropriately chosen quantity of equity which is sold externally. External equity holders are true residual claimants, receiving some (state contingent) fraction of the returns on a firm's assets—less payments to debt-holders. In addition, firm managers (entrepreneurs), whose endowment is their access to investment opportunities, are also residual claimants, being subordinate (in a sense to be made precise) both to debt-holders and outside equity-holders.
What is of greatest interest about this environment, of course, is that it delivers an optimal mix of debt and equity to be used in obtaining external finance. This optimal mix of claims depends on several factors, some of which the firm chooses, and some of which are exogenous to the firm (or the economy). For given parameters, we show that there is a functional relationship between the firm's asset structure and its liability structure. In particular, as the insight provided by the CSV literature would suggest, the use of investment technologies with unobservable returns is associated with debt, while the use of investment technologies with observable returns enables equity finance to be viable. The relationship between the structure of assets and liabilities validates the ancient belief that certain kinds of investments are optimally associated with certain kinds of liability issues.
In terms of factors which are exogenous to the firm, we provide conditions under which the optimal amount of debt finance decreases with monitoring costs. When these costs are sufficiently high, it may be optimal for firms to be financed entirely with equity, and indeed it may be infeasible for them to be financed entirely with debt. Indeed, we state conditions under which 100 percent debt finance is infeasible for a firm; under these conditions a firm which was precluded from issuing liabilities with a sufficient degree of state contingency would be unable to operate. This provides an additional sense in which the ability to issue equity—even in potentially small amounts—can be highly beneficial to a firm.
We also show that factors that are irrelevant to resource allocations are not irrelevant to the liability structure of the firm. For example, with risk neutral agents all resource allocations depend on the distribution of returns on the observable return investment technology only through its mean. However, the debt-equity ratio depends on the entire distribution of returns on this technology. Thus, for example, mean preserving spreads in this distribution are irrelevant to asset allocations, but are far from irrelevant to the composition of firm liabilities. We investigate the relationship between risk and liability structure in some detail.
As should be apparent from this discussion, the environment we describe violates the Modigliani-Miller theorem, and delivers an optimal debt-equity ratio for each firm (or for all firms). Why is it useful to firms to issue one particular mix of debt and equity as opposed to any other? The answer has to do with the nature of investors' responses to the presence of the CSV problem. By assumption, the observable return investment technology has a lower expected return, gross of monitoring costs, than the unobservable return technology does. However, its use permits firms to reduce verification costs. In particular, payments to equity-holders can be conditioned on the returns to investments in the observable return technology. In order to enable the firm to reduce its expected monitoring costs, equity-holders give up resources when the return on the observable return technology is low. In exchange, they receive compensating payments when the return on this technology is high. Moreover, the amount that equity-holders give up (or receive) varies with the outcome of investments in the observable return technology; on average a lot is given up when this return is low, and by way of compensation a lot is received when this return is high. These "transfers" of resources across different return states permit the firm to optimally "smooth" monitoring costs across states. In effect, outside equity-holders allow the firm to use the funds they provide as "collateral" or "internal finance;" the value of both devices in mitigating CSV problems is well known (Bernanke-Gertler 1989 and Boyd-Smith 1994b,c). In this way it is useful to a firm to have a class of external equity claimants, while it also will wish to issue some debt for reasons that are familiar from the standard CSV literature.
The analysis we have just described also provides several ingredients for a sequel (Boyd- Smith 1995), which will examine the evolution of debt and equity market activity as an economy moves along a growth path. In particular, the economy of this paper can be embedded in the neoclassical growth model of Diamond (1965). Under some assumptions about the form taken by monitoring costs, as an economy accumulates capital firms will perceive relative costs of monitoring that change over time. These shifts in perceived (relative) monitoring costs will alter the equilibrium composition of investment, and the equilibrium debt/equity ratio as an economy develops. Boyd and Smith (1995) suggests that the typical expected pattern of development will be that equity markets become increasingly active as an economy grows, and that the composition of investment will become increasingly heavily weighted in favor of the observable return technology. Certainly the former prediction is well-supported by casual observation.
The rest of the study proceeds as follows. Section 1 sets out the model environment. Section 2 describes the financing of investment and defines funding contracts. This section also describes the optimal composition of investment. Section 3 reviews optimal contracts in the standard CSV environment, and contrasts them with optimal contracts in our modified environment. It also shows how an optimal contractual arrangement can be supported by having a firm issue appropriately selected amounts of debt and equity. Section 4 provides some numerical examples indicating how the amount of debt finance (and the debt/equity ratio) depend on the cost of monitoring, and on the risk associated with the various investment technologies. Finally, Section 5 summarizes, offers some conclusions, and relates our results to those obtained in some of the other literature on the optimal composition of firm financial structure.
Download
PDF Ebook The Use of Debt and Equity in: Optimal Financial Contracts
