T he Bond Market Association estimates that U.S. corporations had more than $4.5 trillion in bonds outstanding at the end of 2003, with debt averaging about 50 percent of equity (the value of the stock) from 1994 through 2003. Thus, corporations depend heavily on debt financing. One question that market participants or academic observers have not answered adequately, however, is how companies determine what fraction of their corporate activities should be funded through borrowing and what fraction through issuing stock.

In their seminal 1958 paper, Franco Modigliani and Merton Miller initiated the modern discussion of the amount of debt corporations should use (both received the Nobel Prize for this work and other contributions to economic research). The paper is so well known that, for more than thirty years, financial economists have referred to their theory as “the M&M theory.”

M&M showed that the value of a firm (and of its cash flows) is independent of the ratio of debt to equity used by the firm in financing its investments. This stunning conclusion was based on certain assumptions that are not true of the real world: there are no corporate or personal taxes, people have perfect information. individuals and corporations can borrow at the same rates, and how you pay for an asset does not affect productivity. Still, it provides a jumping-off point for a better understanding of corporate debt.

Here is M&M’s famous arbitrage proof in words. Think of two firms that are identical in all respects, except that one is financed completely with equity while the other uses some combination of equity and debt. Let Ms. E buy 10 percent of the all-equity firm; she buys 10 percent of the outstanding shares. Mr. D buys 10 percent of the leveraged firm; he buys 10 percent of the shares and 10 percent of the debt.

What do Ms. E and Mr. D get back for their investments? In the all-equity firm, Ms. E has a claim on 10 percent of the total profits of the firm. In the leveraged firm, however, the debt holders must receive their interest payments before the shareholders receive the remaining profits. Thus, for his share holdings, Mr. D gets 10 percent of the profits after interest payments to debt holders are subtracted. But because Mr. D also holds 10 percent of the bonds, he receives 10 percent of the profits that were paid out as interest payments. The net result for Mr. D? He receives 10 percent of the total profits, just as Ms. E does.

This reasoning led M&M to argue that the leveraged firm and the all-equity firm must have the exact same value. The value of the all-equity firm is the value of the outstanding stock. The value of the leveraged firm is the value of the outstanding stock plus the value of the outstanding debt. Because the firms are identical in the level of total profits and identical in the cash payouts paid to the investors, Ms. E and Mr. D would pay identical amounts for their respective holdings. M&M went on to show that if the leveraged and all-equity firms do not have the exact same value, arbitragers can make a guaranteed risk-free profit by selling the overvalued firm and buying the undervalued firm.

The proposition that the ratio of debt to equity is irrelevant to the value of the company is known as the irrelevance proposition. Many commentators quickly rejected the irrelevance proposition because its restrictive assumptions did not fit the real world. In 1963, Modigliani and Miller modified their discussion of corporate debt to specifically recognize corporate taxes. Under current tax regulations, interest payments made to bondholders are deducted from corporate income before computation of taxes owed. The corporate tax, then, acts like a subsidy on interest payments. If the corporate tax rate is 34 percent, then for every dollar paid in interest payments, 34 cents in corporate taxes is avoided, though those receiving the interest must pay taxes on it. In contrast, if income is paid out as dividends to shareholders, that income is taxed twice—once at the corporate level and once at the personal level. Therefore, every corporation should minimize its taxes and maximize the cash available to bondand stockholders by financing its investments with 100 percent debt.

This result was more controversial than the first. Firms do not finance their investments with 100 percent debt. Also, there are clear patterns in financing decisions. Young firms in high-growth industries, for example, tend to use less debt, and firms in stable industries with large quantities of fixed assets tend to use more debt. The ensuing study of capital structure and corporate debt has focused on explaining these patterns and explaining why corporations are not 100 percent debt financed.

Financial economists have singled out three additional factors that limit the amount of debt financing: personal taxes, bankruptcy costs, and agency costs. Corporations trade off the benefits of government-subsidized debt against the costs of these three factors. This model of corporate financial structure is therefore called the trade-off theory.

In a 1977 article, Miller showed that considering corporate taxes in isolation is incorrect. Transferring interest payments to individuals to avoid corporate taxes does not make investors any better off if they then have to pay higher personal taxes on that interest income than the corporation and investors would have owed had the corporation not used debt. Miller argued that because tax rates on capital gains have often been lower than tax rates owed on dividend and interest income, the firm might lower the total tax bill paid by the corporation and investor combined by not issuing debt. Moreover, taxes owed on capital gains can be deferred until the realization of those gains, further lowering the effective tax rate on capital gains.

The important thrust of Miller’s argument is that one must look at the interaction of both corporate and personal taxes to determine the optimal level of corporate debt. Miller showed that because of this interaction, there is an optimal level of debt (less than 100 percent) for corporations as a whole. That said, however, he also showed that for any given firm within the economy, the level of debt is again irrelevant as long as the economy-wide average is optimal.

Financial distress or bankruptcy costs may also keep firms from loading up on debt. These financial distress costs take two forms—explicit and implicit. Explicit financial distress costs include the payments made to lawyers, accountants, and so on in filing for Chapter 11 protection from creditors or in liquidation of the firm. These costs can represent a significant portion of corporate assets. Corporations also consider the indirect costs of bankruptcy, including the costs of low inventories, higher costs of inputs from suppliers who fear the company might not pay its bills next month, and the loss of customers who desire a long-term relationship with the firm. The reluctance of travelers to buy airplane tickets from airlines in financial distress or Chapter 11 situations certainly illustrates these indirect costs.

The costs of financial distress are deadweight losses to the investors of the firm: they reduce the cash flows that will eventually be paid to the bondholders and stockholders. Clearly, investors would prefer that firms stay out of financial distress so that these losses are not incurred. As the firm takes on more and more debt, however, the probability of bankruptcy increases. As the amount of debt interest increases, the chance that the firm will default on interest payments increases. These costs restrain firms from maintaining exceptionally high levels of debt.

A third factor limiting the use of debt is “agency costs.” Michael Jensen and William Meckling, in a 1976 article, noted differences between the firm that is 100 percent manager owned and one in which the equity is owned partially by managers and partially by outsiders. In the latter case, the managers act as agents for the outside shareholders. Agents should run the firm to maximize its value. But managers may not be perfect agents, and they may make some decisions in their own interests rather than those of shareholders. Thus, the greater the amount of financing through debt, the greater the concentration of equity in managers’ hands and the less the conflict between managers and shareholders.

However, agency costs also apply to shareholder-bondholder relations. The shareholders, through the managers, have the right to make most decisions about how to run the firm. The firm owes the bondholders fixed payments equal to the amount of money loaned to the firm, along with interest on that principal. Shareholders may adopt policies that benefit themselves at the expense of the bondholders. The possibility for such self-serving behavior is strongest when it is not clear that the firm will have sufficient cash flow to cover its interest and principal loan payments.

The most obvious action shareholders might take to benefit themselves is to pay out all of the firm’s assets as dividends to themselves, leaving an empty shell for the bondholders to claim when the firm is then unable to repay its debt. Shareholders might also follow more subtle strategies. One has been called “risk shifting.” A football analogy illustrates the risk-shifting concept. Woody Hayes, the legendary Ohio State University football coach known for grinding out yardage on the ground, used to say that three things can happen when you pass the ball, and two of them are bad. His philosophy is sound in a close game: it is best to play conservatively and avoid the risk of incompletion or interception. But if you are down by three touchdowns in the fourth quarter, a conservative strategy will almost certainly not win. Instead, you should throw a long pass. True, the ball might be intercepted or fall incomplete, but if you were going to lose anyway, the downside is not that bad. On the upside is the chance of a big payoff—a touchdown.

How does this relate to shareholders and bondholders? If the firm is unlikely to be able to cover its obligations, making the stock worthless, shareholders may throw the long pass—that is, take on risky projects that have big payoffs but high probability of failure. If the project fails, bondholders lose, but the shareholders are no worse off because their claims were worthless anyway. But if the project succeeds, the shareholders will be the major beneficiaries.

A third strategy that may be costly to bondholders is underinvestment on the part of stockholders. If the firm is close to being unable to meet its obligations to bondholders, shareholders may not be willing to put more equity into the firm to fund money-making projects. The reason is that any profits from the new projects are likely to go to bondholders rather than being returned to stockholders. While bondholders would be better off if the projects were undertaken, stockholders will not be willing to pay for them.

All three strategies—paying out large dividends, risk shifting, and underinvestment—are more likely the more indebted is the firm. Lenders know this. Therefore, those who organize the firm, wanting to attract lenders, rationally limit the debt. Bond covenants exist to restrict these games that shareholders might play, but bond contracts cannot prevent all eventualities. An interesting development of the 1980s, however, was the “poison put.” In reaction to the large leveraged buyouts of the 1980s, many companies introduced these poison puts to protect bondholders in the event of a leveraged transaction. Bondholders generally have the right to “put” the bonds to the company and have them repurchased at face value plus, possibly, a small premium if the company takes on a lot of new debt that reduces the chance that the current bondholders will be paid off. These developments illustrate the dynamic nature of corporate finance.

Some researchers have argued that the trade-off theory does not adequately reflect the reality of how managers make financing decisions. An alternative model of financial structure is the “pecking order” model, according to which managers use external financing only when there is insufficient internal financing. In addition, managers prefer to issue debt over equity, if possible, at a reasonable cost of financing. The rationale for the pecking order model is that it is difficult for managers to inform the outside market of the true value of the firm. Thus, there is “asymmetric information” about the value of any securities the firm might issue. Equity, which represents a residual claim on the firm’s assets after all debt holders have been paid, is especially subject to the asymmetric information problem. Since potential investors cannot adequately value stock, it would generally be sold at a price below the price the managers think appropriate. Rather than sell stock too cheaply, therefore, managers who need external financing will prefer to issue debt.

John Graham and Campbell Harvey (2001) surveyed chief financial officers to gather information about their perspective on the determinants of their firms’ financial structure and found support for both the trade-off theory and the pecking order view. They also noted, however, that in many cases, the survey responses were inconsistent with these theories, and they questioned whether the theories were inadequate or if managers were simply ignoring them.

Those attempting to understand the determinants of corporate financing decisions in the United States have looked at empirical regularities among firms for guidance. For example, researchers have found that firms with relatively higher marginal tax burdens are more likely to use debt, thereby taking advantage of the interest tax shield, and firms with more nondebt tax shields, such as depreciation, will use less debt. However, firms that are more likely to face the possibility of financial distress (such as firms with intangible assets or uncertain future cash flows) are more likely to avoid debt in their financing and emphasize equity claims that could not force them into bankruptcy. In addition, firms that may have higher agency costs are also more likely to emphasize equity (see Harris and Raviv 1991).

Interestingly, many of the same regularities in U.S. capital structure exist in other countries. For example, according to Raghuram Rajan and Luigi Zingales (1995), the tax level in a country and the enforcement of bankruptcy laws (and therefore the costs of financial distress) are important determinants of a country’s aggregate debt level. Asli Demirguc-Kunt and Vojislav Maksimovic (1999) showed that firms in countries with stronger legal systems use more external financing and long-term debt than firms in countries with less-developed legal systems.

Researchers continue to investigate the theory of capital structure and the determinants of financing policy around the world. Intercountry comparisons have helped us understand how differences in laws and differences in firms affect firms’ financing policies. An important extension of the global research is investigation of how a country’s financial development influences its economic growth .

About the Author

Annette Poulsen holds the Augustus H. “Billy” Sterne Chair of Banking and Finance at the University of Georgia’s Terry College of Business. She was formerly acting chief economist at the Securities and Exchange Commission.